Instructions to use Undi95/MLewd-L2-13B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Undi95/MLewd-L2-13B-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Undi95/MLewd-L2-13B-GGUF",
	filename="MLewd-L2-13B-v1-8-3.q4_K_S.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Undi95/MLewd-L2-13B-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Undi95/MLewd-L2-13B-GGUF:Q4_K_S
# Run inference directly in the terminal:
llama-cli -hf Undi95/MLewd-L2-13B-GGUF:Q4_K_S

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Undi95/MLewd-L2-13B-GGUF:Q4_K_S
# Run inference directly in the terminal:
llama-cli -hf Undi95/MLewd-L2-13B-GGUF:Q4_K_S

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf Undi95/MLewd-L2-13B-GGUF:Q4_K_S
# Run inference directly in the terminal:
./llama-cli -hf Undi95/MLewd-L2-13B-GGUF:Q4_K_S

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf Undi95/MLewd-L2-13B-GGUF:Q4_K_S
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Undi95/MLewd-L2-13B-GGUF:Q4_K_S

Use Docker

docker model run hf.co/Undi95/MLewd-L2-13B-GGUF:Q4_K_S

LM Studio
Jan
Ollama
How to use Undi95/MLewd-L2-13B-GGUF with Ollama:
```
ollama run hf.co/Undi95/MLewd-L2-13B-GGUF:Q4_K_S
```

Unsloth Studio new

How to use Undi95/MLewd-L2-13B-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Undi95/MLewd-L2-13B-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Undi95/MLewd-L2-13B-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Undi95/MLewd-L2-13B-GGUF to start chatting

Docker Model Runner
How to use Undi95/MLewd-L2-13B-GGUF with Docker Model Runner:
```
docker model run hf.co/Undi95/MLewd-L2-13B-GGUF:Q4_K_S
```

Lemonade

How to use Undi95/MLewd-L2-13B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Undi95/MLewd-L2-13B-GGUF:Q4_K_S

Run and chat with the model

lemonade run user.MLewd-L2-13B-GGUF-Q4_K_S

List all available models

lemonade list

MLewd is a model created to be... Lewd. That's all. Based on ReMM.

There was so much attempt on this model that I can't count them all. Bear with me lmao.

The OG plan: https://pastebin.com/hfJ80rKL

Command useds and explaination :

Due to hardware limitation, some merge was done in 2 part.

Last mix :
- ReMM (Base) (0.57)
- Doctor-Shotgun/llama-2-13b-chat-limarp-v2-merged (Llama Chat Uncensored) (0.35)
- KoboldAI/LLAMA2-13B-Holodeck-1 (0.08)

Part 1: python ties_merge.py TheBloke/Llama-2-13B-fp16 ./MLewdBase-L2-13B-part1  --merge Undi95/ReMM-L2-13B --density 0.88 --merge KoboldAI/LLAMA2-13B-Holodeck-1 --density 0.12 --cuda

Part 2: python ties_merge.py TheBloke/Llama-2-13B-fp16 ./MLewdBase-L2-13B  --merge Undi95/MLewdBase-L2-13B-part1 --density 0.65 --merge Doctor-Shotgun/llama-2-13b-chat-limarp-v2-merged --density 0.35 --cuda

(MLewd-L2-13B-v1-2 got disqualified)

- Applying LoRA: nRuaif/Kimiko-v2-13B at (0.24) weight on MLewd-L2-13B-v1-1
=> Result: MLewd-L2-13B-v1-3

================== ERP RANKING TEST ===========================

19.42 | MLewd-L2-13B-v1-3.q5_K_M.gguf (-> Best)
19.25 | MLewd-L2-13B-v1-1.q5_K_M.gguf
18.25 | MLewd-L2-13B-v1-2.q5 K M.gguf

================== RETRY ===========================

Mix:
- Undi95/MLewd-L2-13B-v1-3 (0.82)
- Sao10K/Stheno-Inverted-L2-13B (0.18)

!python ties_merge.py TheBloke/Llama-2-13B-fp16 ./MLewd-L2-13B-v1-7 --merge Undi95/MLewd-L2-13B-v1-3 --density 0.82 --merge Sao10K/Stheno-Inverted-L2-13B --density 0.18 --cuda
=> Result: MLewd-L2-13B-v1-7


Final touch (trying my best here) :
MLewd-L2-13B-v1-7 (0.77) + zarakiquemparte/PIPPA-ShareGPT-Subset-QLora-13b (LoRA 0.23)
=> MLewd-L2-13B-v1-7-TRY2

FINAL : MLewd-L2-13B-v1-7-TRY2 (0.82) + BluemoonRP (0.18)
=> MLewd-L2-13B-v1-8-3

RIP to all the version that got trashed.

Description

This repo contains quantized files (Q4_K_S and Q5_K_M) of MLewd-L2-13B, a trying-to-be lewd LLM model.

Models used

Undi95/ReMM (Base)
Doctor-Shotgun/llama-2-13b-chat-limarp-v2-merged (Llama Chat Uncensored)
KoboldAI/LLAMA2-13B-Holodeck-1
Sao10K/Stheno-Inverted-L2-13B

Loras used

nRuaif/BluemoonRP-L2-13B-This-time-will-be-better/tree/main/lora-out-13b-final-BM/checkpoint-15/adapter_model
zarakiquemparte/PIPPA-ShareGPT-Subset-QLora-13b

Prompt template: Alpaca

Below is an instruction that describes a task. Write a response that appropriately completes the request.

### Instruction:
{prompt}

### Response:

Special thanks to Sushi kek

Downloads last month: 14

GGUF

Model size

13B params

Architecture

llama

Hardware compatibility

4-bit

5-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support