Instructions to use zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw with Transformers:

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

pipe = pipeline("text-generation", model="zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw")
model = AutoModelForCausalLM.from_pretrained("zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw", 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 zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw

SGLang

How to use zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw 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 "zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw" \
    --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": "zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw",
		"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 "zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw" \
        --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": "zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw with Docker Model Runner:
```
docker model run hf.co/zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw
```

PAINTED FANTASY

Mistral Small 3.2 24B

Overview

Experimental release.

This is an uncensored creative model intended to excel at character driven RP / ERP.

This model is designed to provide longer, narrative heavy responses where characters are portrayed accurately and proactively.

SillyTavern Settings

Recommended Roleplay Format

> Actions: In plaintext

> Dialogue: "In quotes"

> Thoughts: *In asterisks*

Recommended Samplers

> Temp: 0.8

> MinP: 0.04 - 0.05

> TopP: 0.95 - 1.0

> Dry: 0.8, 1.75, 4

Instruct

Mistral v7 Tekken

Quantizations

GGUF

> Static (mrademacher)

> iMatrix (mrademacher)

EXL3

> 3bpw

> 4bpw

> 5bpw

> 6bpw

Training Process

Training process: Pretrain > SFT > DPO > DPO 2

Did a small pretrain on some light novels and Frieren wiki data as a test. Hasn't seemed to hurt the model and model has shown some small improvements in the lore of series that were included.

The model then went through the standard SFT using a dataset of approx 3.6 million tokens, 700 RP conversations, 1000 creative writing / instruct samples and about 100 summaries. The bulk of this data has been made public.

Finally DPO was used to make the model a little more consistent. The first stage of DPO focused on instruction following and the second tried to burn out some Mistral-isms.

Not optimized for cost / performance efficiency, YMMV.

SFT 1*H100

# ====================
# MODEL CONFIGURATION
# ====================
base_model: ./MS3-2-Pretrain/merged
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer


# ====================
# DATASET CONFIGURATION
# ====================
datasets:
  - path: ./dataset.jsonl
    type: chat_template
    split: train
    chat_template_strategy: tokenizer
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]

dataset_prepared_path:
train_on_inputs: false  # Only train on assistant responses
# ====================
# QLORA CONFIGURATION
# ====================
adapter: qlora
load_in_4bit: true
lora_r: 128
lora_alpha: 128
lora_dropout: 0.1
lora_target_linear: true
# lora_modules_to_save:  # Uncomment only if you added NEW tokens
# ====================
# TRAINING PARAMETERS
# ====================
num_epochs: 3
micro_batch_size: 4
gradient_accumulation_steps: 2
learning_rate: 1e-5
optimizer: paged_adamw_8bit
lr_scheduler: rex
warmup_ratio: 0.05
weight_decay: 0.01
max_grad_norm: 1.0
# ====================
# SEQUENCE & PACKING
# ====================
sequence_len: 8192
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true
# ====================
# HARDWARE OPTIMIZATIONS
# ====================
bf16: auto
flash_attention: true
gradient_checkpointing: true
# ====================
# EVALUATION & CHECKPOINTING
# ====================
save_strategy: steps
save_steps: 5
save_total_limit: 5  # Keep best + last few checkpoints
load_best_model_at_end: true
greater_is_better: false
# ====================
# LOGGING & OUTPUT
# ====================
output_dir: ./MS3-2-SFT-2
logging_steps: 2
save_safetensors: true
# ====================
# WANDB TRACKING
# ====================
wandb_project: MS3-2-SFT
wandb_entity: your_entity
wandb_name: run_name

Downloads last month: 7

Safetensors

Model size

6B params

Tensor type

F16

I16

Model tree for zerofata/MS3.2-PaintedFantasy-24B_exl3-4bpw

Base model

mistralai/Mistral-Small-3.1-24B-Base-2503

Finetuned

mistralai/Mistral-Small-3.2-24B-Instruct-2506

Finetuned

zerofata/MS3.2-PaintedFantasy-24B

Quantized

(6)

this model