Instructions to use ewald1976/findesiecle-12b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ewald1976/findesiecle-12b with Transformers:

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

pipe = pipeline("text-generation", model="ewald1976/findesiecle-12b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("ewald1976/findesiecle-12b")
model = AutoModelForCausalLM.from_pretrained("ewald1976/findesiecle-12b")
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]:]))

llama-cpp-python

How to use ewald1976/findesiecle-12b with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="ewald1976/findesiecle-12b",
	filename="findesiecle-12b-Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Inference
Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use ewald1976/findesiecle-12b with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf ewald1976/findesiecle-12b:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf ewald1976/findesiecle-12b:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf ewald1976/findesiecle-12b:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf ewald1976/findesiecle-12b:Q4_K_M

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 ewald1976/findesiecle-12b:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf ewald1976/findesiecle-12b:Q4_K_M

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 ewald1976/findesiecle-12b:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ewald1976/findesiecle-12b:Q4_K_M

Use Docker

docker model run hf.co/ewald1976/findesiecle-12b:Q4_K_M

LM Studio
Jan

vLLM

How to use ewald1976/findesiecle-12b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ewald1976/findesiecle-12b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ewald1976/findesiecle-12b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ewald1976/findesiecle-12b:Q4_K_M

SGLang

How to use ewald1976/findesiecle-12b 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 "ewald1976/findesiecle-12b" \
    --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": "ewald1976/findesiecle-12b",
		"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 "ewald1976/findesiecle-12b" \
        --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": "ewald1976/findesiecle-12b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use ewald1976/findesiecle-12b with Ollama:
```
ollama run hf.co/ewald1976/findesiecle-12b:Q4_K_M
```

Unsloth Studio new

How to use ewald1976/findesiecle-12b 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 ewald1976/findesiecle-12b 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 ewald1976/findesiecle-12b to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for ewald1976/findesiecle-12b to start chatting

Pi new

How to use ewald1976/findesiecle-12b with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf ewald1976/findesiecle-12b:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "ewald1976/findesiecle-12b:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use ewald1976/findesiecle-12b with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf ewald1976/findesiecle-12b:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default ewald1976/findesiecle-12b:Q4_K_M

Run Hermes

hermes

Docker Model Runner
How to use ewald1976/findesiecle-12b with Docker Model Runner:
```
docker model run hf.co/ewald1976/findesiecle-12b:Q4_K_M
```

Lemonade

How to use ewald1976/findesiecle-12b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ewald1976/findesiecle-12b:Q4_K_M

Run and chat with the model

lemonade run user.findesiecle-12b-Q4_K_M

List all available models

lemonade list

findesiecle-12b

A creative writing fine-tune of Mistral Nemo 12B trained on a curated literary dataset in the English contemplative tradition.

What this model is

findesiecle-12b is intended for roleplay, creative writing, and narrative generation that values stylistic restraint over dramatic intensity. It has been shaped by writers such as Huxley, Wells, Shelley, Machen, Blackwood, Stapledon, and the broader Edwardian and inter-war English literary tradition — extended through the contemplative sensibility one might associate with the phrase quiet desperation.

It writes, in short, the way a thoughtful person reads.

What this model is not

This model is not what most current 12B roleplay fine-tunes are.

If you are looking for:

explicit NSFW content
dark/edgy/transgressive output as a feature
shock value, gore, or graphic violence
aggressive uncensored "based" output
erotic roleplay as a primary use case

— this is probably not the model for you. There are many excellent fine-tunes that cater to those use cases. findesiecle-12b is not in competition with them; it serves a different purpose, for a different reader.

The dataset contains no NSFW material. The model has not been trained to produce it, and while it is technically uncensored, it will not naturally lean in that direction. The literary tradition it draws from treats sensuality, when it appears at all, with restraint and implication rather than explicit description.

Who this model is for

If you have ever:

finished an 800-token roleplay output and felt your time was wasted
wished a model could write a scene in which nothing dramatic happens and yet something is conveyed
wanted to roleplay a character whose interior life is more interesting than their action
found yourself nostalgic for prose that trusts the reader

— this model may be worth your time.

Style characteristics

findesiecle-12b tends toward:

economy — sentences carry weight; adjective chains are avoided
observation over explanation — what is seen rather than what it means
sachlichkeit before the extraordinary — uncanny or remarkable events are described more, not less, calmly
space between sentences — silences are part of the prose
period sensibility — without modernizing reflex; comfortable in the 19th and early 20th centuries
psychological subtext — characters reveal themselves through what they do not say

Sample output

Prompt: Describe an old woman writing a letter to a sister she has not seen in thirty years.

Dear Margaret,

I have been sitting at this desk for some considerable time. The lamp is lit, although it is not yet four. The room is warm enough. I have been holding the pen without writing.

I shall begin.

The house is much as it was. Mother's chair remains by the window, although the curtains have been replaced in the intervening years. Father's books are on the lower shelf of the bookcase; I have not, in thirty years, added to them.

[...]

Margaret, I have been a good sister, by the only measure available to me. I have not written. I have not visited. I have kept the small distance which our lives have, by long habit, maintained. I have not, in thirty years, offered any reason for this distance. I have not, in thirty years, been asked for one.

I shall not, in this letter, offer a reason. I shall not, in this letter, ask for one.

Recommended inference settings

Lower temperature than usual works well — the model's strength is its restraint, and high temperature can produce uncharacteristic verbosity.

Temperature: 0.6 – 0.8
Top-p: 0.9
Repetition penalty: 1.05 – 1.10

A note on intention

This is the first in a planned series. The fine-tune exists because the maintainer found themselves unable to read the output of most current 12B roleplay models with any pleasure, and decided to build the model they wished existed.

It is offered to others who may share that condition.