Instructions to use AaryanK/NousCoder-14B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AaryanK/NousCoder-14B-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="AaryanK/NousCoder-14B-GGUF",
	filename="NousCoder-14B.q2_k.gguf",
)

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

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use AaryanK/NousCoder-14B-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf AaryanK/NousCoder-14B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf AaryanK/NousCoder-14B-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf AaryanK/NousCoder-14B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf AaryanK/NousCoder-14B-GGUF: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 AaryanK/NousCoder-14B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf AaryanK/NousCoder-14B-GGUF: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 AaryanK/NousCoder-14B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf AaryanK/NousCoder-14B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/AaryanK/NousCoder-14B-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use AaryanK/NousCoder-14B-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AaryanK/NousCoder-14B-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AaryanK/NousCoder-14B-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AaryanK/NousCoder-14B-GGUF:Q4_K_M

Ollama
How to use AaryanK/NousCoder-14B-GGUF with Ollama:
```
ollama run hf.co/AaryanK/NousCoder-14B-GGUF:Q4_K_M
```

Unsloth Studio new

How to use AaryanK/NousCoder-14B-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 AaryanK/NousCoder-14B-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 AaryanK/NousCoder-14B-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for AaryanK/NousCoder-14B-GGUF to start chatting

Pi new

How to use AaryanK/NousCoder-14B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf AaryanK/NousCoder-14B-GGUF: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": "AaryanK/NousCoder-14B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use AaryanK/NousCoder-14B-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf AaryanK/NousCoder-14B-GGUF: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 AaryanK/NousCoder-14B-GGUF:Q4_K_M

Run Hermes

hermes

Docker Model Runner
How to use AaryanK/NousCoder-14B-GGUF with Docker Model Runner:
```
docker model run hf.co/AaryanK/NousCoder-14B-GGUF:Q4_K_M
```

Lemonade

How to use AaryanK/NousCoder-14B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull AaryanK/NousCoder-14B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.NousCoder-14B-GGUF-Q4_K_M

List all available models

lemonade list

NousCoder-14B-GGUF

GGUF format model files for NousCoder-14B.

Available Files

Filename	Quant Type	Size	Description
`NousCoder-14B.q2_k.gguf`	Q2_K	5.36 GB	Smallest size, lowest quality. Best for testing if model works.
`NousCoder-14B.q3_k_l.gguf`	Q3_K_L (Large)	7.36 GB	Better quality 3-bit quantization.
`NousCoder-14B.q3_k_m.gguf`	Q3_K_M (Medium)	6.82 GB	Small size with slightly better quality than Q3_K_S.
`NousCoder-14B.q3_k_s.gguf`	Q3_K_S (Small)	6.20 GB	Very small, low quality. Not recommended for most uses.
`NousCoder-14B.q4_0.gguf`	Q4_0	7.93 GB	Basic 4-bit quantization. Good balance of size and quality.
`NousCoder-14B.q4_1.gguf`	Q4_1	8.74 GB	4-bit with higher accuracy than Q4_0.
`NousCoder-14B.q4_k_m.gguf`	Q4_K_M (Medium)	8.38 GB	K-quant 4-bit, best balance of size and quality. Most popular choice.
`NousCoder-14B.q4_k_s.gguf`	Q4_K_S (Small)	7.98 GB	K-quant 4-bit, optimized for smaller size.
`NousCoder-14B.q5_0.gguf`	Q5_0	9.56 GB	Basic 5-bit quantization. Higher quality than Q4.
`NousCoder-14B.q5_1.gguf`	Q5_1	10.37 GB	5-bit with higher accuracy than Q5_0.
`NousCoder-14B.q5_k_m.gguf`	Q5_K_M (Medium)	9.79 GB	K-quant 5-bit, excellent quality-to-size ratio.
`NousCoder-14B.q5_k_s.gguf`	Q5_K_S (Small)	9.56 GB	K-quant 5-bit, good quality with reasonable size.
`NousCoder-14B.q6_k.gguf`	Q6_K	11.29 GB	6-bit quantization. Very high quality, larger size.
`NousCoder-14B.q8_0.gguf`	Q8_0	14.62 GB	8-bit quantization. Near-original quality, largest quantized size.

Usage

# With llama.cpp
./llama-cli -m NousCoder-14B.q4_k_m.gguf -p "Your prompt" -n 128

Original Model Card

NousCoder-14B

We introduce NousCoder-14B, a competitive programming model post-trained on Qwen3-14B via reinforcement learning. On LiveCodeBench v6 (08/01/2024 - 05/01/2025), we achieve a Pass@1 accuracy of 67.87%, up 7.08% from the baseline Pass@1 accuracy of 60.79% of Qwen3-14B. We trained on 24k verifiable coding problems using 48 B200s over the course of four days.

Acknowledgements

I would like to thank my mentor, Roger Jin, Dakota Mahan, Teknium, and others at the Nous Research team for their invaluable support throughout this project. I would also like to thank Together AI and Agentica for their immensely helpful blog posts on DeepCoder-14B. Finally, thank you to Modal and Lambda for their generous support by providing me with credits.

Downloads last month: 98

GGUF

Model size

15B params

Architecture

qwen3

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

Model tree for AaryanK/NousCoder-14B-GGUF

Base model

Qwen/Qwen3-14B-Base

Finetuned

Qwen/Qwen3-14B

Finetuned

NousResearch/NousCoder-14B

Quantized

(20)

this model