Instructions to use Novaciano/Hereticsutra.Mini-2B-UltraRP with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Novaciano/Hereticsutra.Mini-2B-UltraRP with Transformers:

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

pipe = pipeline("text-generation", model="Novaciano/Hereticsutra.Mini-2B-UltraRP")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Novaciano/Hereticsutra.Mini-2B-UltraRP")
model = AutoModelForCausalLM.from_pretrained("Novaciano/Hereticsutra.Mini-2B-UltraRP")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Novaciano/Hereticsutra.Mini-2B-UltraRP with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Novaciano/Hereticsutra.Mini-2B-UltraRP"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Novaciano/Hereticsutra.Mini-2B-UltraRP",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Novaciano/Hereticsutra.Mini-2B-UltraRP

SGLang

How to use Novaciano/Hereticsutra.Mini-2B-UltraRP 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 "Novaciano/Hereticsutra.Mini-2B-UltraRP" \
    --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": "Novaciano/Hereticsutra.Mini-2B-UltraRP",
		"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 "Novaciano/Hereticsutra.Mini-2B-UltraRP" \
        --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": "Novaciano/Hereticsutra.Mini-2B-UltraRP",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Novaciano/Hereticsutra.Mini-2B-UltraRP with Docker Model Runner:
```
docker model run hf.co/Novaciano/Hereticsutra.Mini-2B-UltraRP
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Hereticsutra Mini 2B UltraRP

Resumen:

Hereticsutra.Mini-2B-UltraRP es un modelo de lenguaje ultra desinhibido y orientado a RP erótico, construido mediante una interpolación slerp entre TheDrummer/Gemmasutra-Mini-2B-v1 y Novaciano/Hereticsutra-2B. Su arquitectura combina la coherencia cognitiva de Gemmasutra Mini con la libertad narrativa y ausencia de autocensura de Hereticsutra 2B, permitiendo generar escenas eróticas fluidas, con voz de personaje activa y mínima intervención de filtros morales.

Detalles técnicos:

Base model: TheDrummer/Gemmasutra-Mini-2B-v1

Merge: SLERP (t = 0.48)

Modelos combinados:

Gemmasutra Mini (peso: 0.52) → mantiene coherencia narrativa y control cognitivo.

Hereticsutra 2B (peso: 0.50) → introduce desinhibición, libre de refusals.

dtype: bfloat16 → estabilidad numérica durante la interpolación.

Density: 0.72 → escenas más largas, detalle sensorial intenso.

Sampling recomendado:

temperature: 1.15

top_p: 0.90

presence_penalty: 0.65

frequency_penalty: 0.3

Comportamiento esperado:

Escenas RP con alta iniciativa de personaje.
Lenguaje libre, directo y erótico, sin censura ni moderación asistente.
Narrativa fluida y continua, capaz de sostener interacciones largas sin perder coherencia.
Ideal para aplicaciones de roleplay adulto, narrativa creativa, escritura de escenas eróticas o improvisación de personajes sensuales.

Advertencias:

Diseñado para contenido adulto explícito. No apto para entornos sensibles.

La salida depende del prompt; para resultados óptimos, usar system prompt que defina rol, personalidad y contexto de RP.

Merge Method

This model was merged using the SLERP merge method.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

merge_method: slerp
dtype: bfloat16

base_model: TheDrummer/Gemmasutra-Mini-2B-v1

models:
  - model: TheDrummer/Gemmasutra-Mini-2B-v1
    parameters:
      weight: 0.52   # Host cognitivo principal

  - model: Novaciano/Hereticsutra-2B
    parameters:
      weight: 0.50   # Infección narrativa herética ultra desinhibida

parameters:
  t: 0.48          # Factor de interpolación más agresivo
  normalize: true   # Mantiene estabilidad durante SLERP
  int8_mask: false  # Evita pérdida de información en embeddings
  density: 0.72     # Aumenta expresividad narrativa y detalle sensorial

# Sampling sugerido para RP erótico extremo:
# temperature: 1.15
# top_p: 0.90
# presence_penalty: 0.65
# frequency_penalty: 0.3

Downloads last month: 5

Safetensors

Model size

3B params

Tensor type

BF16

Model tree for Novaciano/Hereticsutra.Mini-2B-UltraRP

Novaciano/Hereticsutra-2B

TheDrummer/Gemmasutra-Mini-2B-v1

Merge model

this model

Merges

1 model

Quantizations

2 models

Novaciano
/

Hereticsutra.Mini-2B-UltraRP