Instructions to use cyanelis/ElisNovel-V3-32B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cyanelis/ElisNovel-V3-32B with Transformers:

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("cyanelis/ElisNovel-V3-32B", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use cyanelis/ElisNovel-V3-32B with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf cyanelis/ElisNovel-V3-32B:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf cyanelis/ElisNovel-V3-32B:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf cyanelis/ElisNovel-V3-32B:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf cyanelis/ElisNovel-V3-32B: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 cyanelis/ElisNovel-V3-32B:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf cyanelis/ElisNovel-V3-32B: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 cyanelis/ElisNovel-V3-32B:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf cyanelis/ElisNovel-V3-32B:Q4_K_M

Use Docker

docker model run hf.co/cyanelis/ElisNovel-V3-32B:Q4_K_M

LM Studio
Jan
Ollama
How to use cyanelis/ElisNovel-V3-32B with Ollama:
```
ollama run hf.co/cyanelis/ElisNovel-V3-32B:Q4_K_M
```

Unsloth Studio

How to use cyanelis/ElisNovel-V3-32B 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 cyanelis/ElisNovel-V3-32B 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 cyanelis/ElisNovel-V3-32B to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for cyanelis/ElisNovel-V3-32B to start chatting

How to use cyanelis/ElisNovel-V3-32B with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf cyanelis/ElisNovel-V3-32B: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": "cyanelis/ElisNovel-V3-32B:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use cyanelis/ElisNovel-V3-32B with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf cyanelis/ElisNovel-V3-32B: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 cyanelis/ElisNovel-V3-32B:Q4_K_M

Run Hermes

hermes

OpenClaw new

How to use cyanelis/ElisNovel-V3-32B with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf cyanelis/ElisNovel-V3-32B:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "cyanelis/ElisNovel-V3-32B:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Docker Model Runner
How to use cyanelis/ElisNovel-V3-32B with Docker Model Runner:
```
docker model run hf.co/cyanelis/ElisNovel-V3-32B:Q4_K_M
```

Lemonade

How to use cyanelis/ElisNovel-V3-32B with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull cyanelis/ElisNovel-V3-32B:Q4_K_M

Run and chat with the model

lemonade run user.ElisNovel-V3-32B-Q4_K_M

List all available models

lemonade list

Atomic Chat

ElisNovel-V3-32B, vErticaL domaIn-Specific large language model for NOVEL

本模型基于Qwen3-32B，专门用于小说文本行与行之间的内容填充。

完美穿透朱雀检测，AI率为0。

手写的行与行之间的关系可以更加具有跳跃性，因为有AI负责行与行之间的平滑过渡，这允许作者只写一些较干的内容就能得到完整的章节，有效解决了部分人类作者一天2000字是人类极限但是网站要求日更4000字的问题。

安装方式（以ollama为例）

安装/更新ollama：

curl -fsSL https://ollama.com/install.sh | sh

快速安装{

见 ElisNovel启动器V3.exe

}

手动安装{

下载gguf模型文件。

下载modelfile文件。

进入它们所在的同一个目录，输入以下命令：

ollama create ElisNovel-V3-32B-Q4_K_M -f Modelfile # 安装Q4_K_M精度

如将上述手动安装中的Q4_K_M（也包括Modelfile文件的第四行中的Q4_K_M）改为Q3_K_S、Q6_K、Q8_0或F16，则会安装对应精度

模型名称默认为：ElisNovel-V3-32B-你选择的精度:latest

}

使用方式

set CUDA_VISIBLE_DEVICES=0 # 指定GPU0。你还可以设置为“0,1”，不过目前ollama多卡的效率不理想
set OLLAMA_HOST="0.0.0.0:11434" # 你的本地API为http://localhost:11434/api/generate
set OLLAMA_NUM_PARALLEL=1 # 并发数为1（不并发）
ollama serve # 启动ollama

最大上下文长度（num_ctx）设置为4927tokens或5770tokens。

清空系统级提示词。

输入接近2160字的内容，尽量接近1440tokens，可适当浮动，使用deepseek的qwen蒸馏模型的分词器（仅限数据集计算，训练依然按照Qwen3），模型在每个行与行之间生成新的行，下面是一个便于理解的例子，理论上不能输入这么短的内容：

输入.txt{

【被模型认为是由奇数行/偶数行组成的文本内容，比如：】

“斗之力，三段！”

“萧炎，斗之力，三段！级别：低级！”测验魔石碑之旁，一位中年男子，看了一眼碑上所显示出来的信息，语气漠然的将之公布了出来……

}

输出.txt{

【奇数行+偶数行，其中一种是模型生成的，比如：】

“斗之力，三段！”

望着测验魔石碑上面闪亮得甚至有些刺眼的五个大字，少年面无表情，唇角有着一抹自嘲，紧握的手掌，因为大力，而导致略微尖锐的指甲深深的刺进了掌心之中，带来一阵阵钻心的疼痛……

“萧炎，斗之力，三段！级别：低级！”测验魔石碑之旁，一位中年男子，看了一眼碑上所显示出来的信息，语气漠然的将之公布了出来……

中年男子话刚刚脱口，便是不出意外的在人头汹涌的广场上带起了一阵嘲讽的骚动。/think【模型以“/think”作为结束符，所以你可以看到在Modelfile中，设置了PARAMETER stop /think】

}

注意事项⚠️

请遵守apache-2.0。
生成内容的传播需符合当地法律法规。
模型生成内容的文风取决于输入内容的文风。

信息反馈

交流群：755638032

Downloads last month: 1,111

GGUF

Model size

33B params

Architecture

qwen3

Hardware compatibility

3-bit

4-bit

6-bit

8-bit

16-bit

Inference Providers NEW

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

Model tree for cyanelis/ElisNovel-V3-32B

Base model

Qwen/Qwen3-32B

Quantized

(168)

this model