Instructions to use OccultAI/MN-Adversary-12B-v1.62 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use OccultAI/MN-Adversary-12B-v1.62 with Transformers:

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

pipe = pipeline("text-generation", model="OccultAI/MN-Adversary-12B-v1.62")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("OccultAI/MN-Adversary-12B-v1.62")
model = AutoModelForCausalLM.from_pretrained("OccultAI/MN-Adversary-12B-v1.62", device_map="auto")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use OccultAI/MN-Adversary-12B-v1.62 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "OccultAI/MN-Adversary-12B-v1.62"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "OccultAI/MN-Adversary-12B-v1.62",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/OccultAI/MN-Adversary-12B-v1.62

SGLang

How to use OccultAI/MN-Adversary-12B-v1.62 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 "OccultAI/MN-Adversary-12B-v1.62" \
    --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": "OccultAI/MN-Adversary-12B-v1.62",
		"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 "OccultAI/MN-Adversary-12B-v1.62" \
        --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": "OccultAI/MN-Adversary-12B-v1.62",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use OccultAI/MN-Adversary-12B-v1.62 with Docker Model Runner:
```
docker model run hf.co/OccultAI/MN-Adversary-12B-v1.62
```

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.

⚠️ Warning: This model can produce narratives and RP that contain violent and graphic erotic content. Use ChatML template for best results, although Mistral Tekken works sometimes.

👹 Adversary 12B v1.62

Fully uncensored, evil, and creative. Trained in the dark arts of black magick and Ars Goetia.

This is a merge of pre-trained language models created using mergekit.

Included in the merge are 3 versions of the Adversary finetune, as well as the merged Raven 12B v1 finetunes, and an unreleased Morpheus 12B v1 finetune.

This model was merged using the DELLA merge method as well as the MULTI_FUSION merge method.

Runpod was used to help generate the datasets and LoRAs, though merges were done locally.

Recommended Temp

0.3

Models Merged

The following models were included in the merge:

SicariusSicariiStuff/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct_Abliterated
DarkArtsForge/MN-Adversary-12B-v1.62
DarkArtsForge/MN-Raven-12B-v1
DarkArtsForge/MN-Morpheus-12B-v1_epoch2
Lambent/Arsenic-Shahrazad-12B-v4.4
Retreatcost/Evertide-RX-12B
KOOWEEYUS/BlackSheep-RP-12B
WokeAI/Tankie-DPE-12B-SFT-v2
DavidAU/Mistral-Nemo-2407-12B-Thinking-Claude-Gemini-GPT5.2-Uncensored-HERETIC
DavidAU-/MN-Dark-Planet-TITAN-12B

Configuration

The following YAML configurations were used to produce this model:

architecture: MistralForCausalLM
base_model: B:\12B\MN-Adversary-12B-v1.44-r64
models:
  - model: B:\12B\MN-Adversary-12B-v1.44-r64 # lora
  - model: B:\12B\MN-Adversary-12B-v1.45 # lora
merge_method: multi_fusion
parameters:  
  tukey_fence: 1.5
  importance_metric: "delta_mag"
dtype: float32
out_dtype: bfloat16
tokenizer:
  source: union
chat_template: auto
name: MN-Adversary-12B-v1.46

architecture: MistralForCausalLM
base_model: B:\12B\MuXodious--Mistral-Nemo-Instruct-2407-absolute-heresy
models:
  - model: B:\12B\MN-Raven-12B-v1 # lora
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: B:\12B\MN-Adversary-12B-v1.5 # lora
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: B:\12B\MN-Adversary-12B-v1.46 # merge
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: B:\12B\MN-Morpheus-12B-v1-Base-LoRA\MN-Morpheus-12B-v1_epoch2
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: B:\12B\Lambent--Arsenic-Shahrazad-12B-v4.4
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: B:\12B\Retreatcost--Evertide-RX-12B
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: A:\LLM\.cache\13B\KOOWEEYUS--BlackSheep-RP-12B
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: A:\LLM\.cache\13B\WokeAI--Tankie-DPE-12B-SFT-v2
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: B:\12B\DavidAU--Mistral-Nemo-2407-12B-Thinking-Claude-Gemini-GPT5.2-Uncensored-HERETIC
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
  - model: B:\12B\DavidAU--MN-Dark-Planet-TITAN-12B
    parameters:
      weight: 0.2
      density: 0.9
      epsilon: 0.09
merge_method: della
parameters:
  lambda: 1.0
  normalize: false
  int8_mask: false
  rescale: true
dtype: float32
out_dtype: bfloat16
tokenizer:
  source: union
name: MN-Adversary-12B-v1.62