nsfw

Merge

mergekit

conversational

text-generation-inference

Model card Files Files and versions

xet

Community

Instructions to use Bumoch/Forgotten_Magic_24B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

Transformers

How to use Bumoch/Forgotten_Magic_24B with Transformers:

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

pipe = pipeline("text-generation", model="Bumoch/Forgotten_Magic_24B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Bumoch/Forgotten_Magic_24B")
model = AutoModelForCausalLM.from_pretrained("Bumoch/Forgotten_Magic_24B", 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 Bumoch/Forgotten_Magic_24B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Bumoch/Forgotten_Magic_24B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Bumoch/Forgotten_Magic_24B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Bumoch/Forgotten_Magic_24B

SGLang

How to use Bumoch/Forgotten_Magic_24B 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 "Bumoch/Forgotten_Magic_24B" \
    --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": "Bumoch/Forgotten_Magic_24B",
		"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 "Bumoch/Forgotten_Magic_24B" \
        --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": "Bumoch/Forgotten_Magic_24B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Bumoch/Forgotten_Magic_24B with Docker Model Runner:
```
docker model run hf.co/Bumoch/Forgotten_Magic_24B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

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.

ABANDONED !!!

Merge Method

This model was merged using the Karcher Mean merge method.

Quants (THX to Mradermacher and Group!)

https://hf.tst.eu/model#Forgotten_Magic_24B-GGUF

Models Merged

Magistry-24B-v1.1
Asmodeus-24B-v2
Cydonia-24B-v4.3-absolute-heresy
Forgotten-Abomination-24B-V3.0
Forgotten-Safeword-24B-3.4
Magidonia-24B-v4.3
Precog-24B-v1

first impression so far

The model is clever, even with over 20K tokens. I haven't tested it fully yet, but so far it's pleasant to converse with. It doesn't choose to be NSFW directly, but it evaluates and analyses the situation. I hope I will get some feedback to help me decide whether to fine-tune it a little, abandon it, or leave it as it is. In terms of impersonating the user, it only did so twice, but when I removed parts of the reply, it actually went well.

Update

Yes at times it wants to speak and act as the user, but if deleted or re done, it works. The merge is amazingly clever. If is not aiming for intimacy or gore, but also not avoiding it, it is kinda interesting and needs more testing.

Downloads last month: 2

Safetensors

Model size

24B params

Tensor type

BF16

Model tree for Bumoch/Forgotten_Magic_24B

DarkArtsForge/Asmodeus-24B-v2

MuXodious/Cydonia-24B-v4.3-absolute-heresy

ReadyArt/Forgotten-Abomination-24B-V3.0

ReadyArt/Forgotten-Safeword-24B-3.4

TheDrummer/Magidonia-24B-v4.3

TheDrummer/Precog-24B-v1

sophosympatheia/Magistry-24B-v1.1

Merge model

this model

Quantizations

2 models