Llama 3.1 Nemotron Nano 4B V1.1 GGUF

🚀 Llama-3.1-Nemotron-Nano-4B-v1.1 GGUF Models

This project provides the Llama-3.1-Nemotron-Nano-4B-v1.1 GGUF models, offering a balance between accuracy and efficiency for various text - generation tasks. It includes different quantization methods and model formats to suit diverse hardware and memory requirements.

✨ Features

• Ultra - Low - Bit Quantization: Introduces precision - adaptive quantization for 1 - 2 bit models, improving accuracy on Llama - 3 - 8B.
• Multiple Model Formats: Offers BF16, F16, and various quantized formats to fit different hardware capabilities and memory constraints.
• Tool - Call Support: Supports tool calling and can be used to launch a vLLM server.

📦 Installation

The README does not provide specific installation steps, so this section is skipped.

💻 Usage Examples

Basic Usage

import torch
import transformers

model_id = "nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1"
model_kwargs = {"torch_dtype": torch.bfloat16, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id

pipeline = transformers.pipeline(
   "text-generation",
   model=model_id,
   tokenizer=tokenizer,
   max_new_tokens=32768,
   temperature=0.6,
   top_p=0.95,
   **model_kwargs
)

# Thinking can be "on" or "off"
thinking = "on"

print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"}, {"role": "user", "content": "Solve x*(sin(x)+2)=0"}]))

Advanced Usage

import torch
import transformers

model_id = "nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1"
model_kwargs = {"torch_dtype": torch.bfloat16, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id

# Thinking can be "on" or "off"
thinking = "off"

pipeline = transformers.pipeline(
   "text-generation",
   model=model_id,
   tokenizer=tokenizer,
   max_new_tokens=32768,
   do_sample=False,
   **model_kwargs
)

print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"}, {"role": "user", "content": "Solve x*(sin(x)+2)=0"}, {"role":"assistant", "content":"<think>\n</think>"}]))

📚 Documentation

Model Generation Details

This model was generated using llama.cpp at commit 92ecdcc0.

Ultra - Low - Bit Quantization with IQ - DynamicGate (1 - 2 bit)

Our latest quantization method introduces precision - adaptive quantization for ultra - low - bit models (1 - 2 bit), with benchmark - proven improvements on Llama - 3 - 8B. This approach uses layer - specific strategies to preserve accuracy while maintaining extreme memory efficiency.

Benchmark Context

All tests were conducted on Llama - 3 - 8B - Instruct using:

• Standard perplexity evaluation pipeline
• 2048 - token context window
• The same prompt set across all quantizations

Method

• Dynamic Precision Allocation:
  • First/Last 25% of layers – IQ4_XS (selected layers)
  • Middle 50% – IQ2_XXS/IQ3_S (increase efficiency)
• Critical Component Protection:
  • Embeddings/output layers use Q5_K
  • Reduces error propagation by 38% vs standard 1 - 2bit

Quantization Performance Comparison (Llama - 3 - 8B)

Key:

• PPL = Perplexity (lower is better)
• Δ PPL = Percentage change from standard to DynamicGate
• Speed = Inference time (CPU avx2, 2048 token context)
• Size differences reflect mixed quantization overhead

Key Improvements:

• IQ1_M shows a massive 43.9% perplexity reduction (27.46 – 15.41)
• IQ2_S cuts perplexity by 36.9% while adding only 0.2GB
• IQ1_S maintains 39.7% better accuracy despite 1 - bit quantization

Tradeoffs:

• All variants have modest size increases (0.1 - 0.3GB)
• Inference speeds remain comparable (<5% difference)

When to Use These Models

• Fitting models into GPU VRAM
• Memory - constrained deployments
• CPU and Edge Devices where 1 - 2bit errors can be tolerated
• Research into ultra - low - bit quantization

Choosing the Right Model Format

Selecting the correct model format depends on your hardware capabilities and memory constraints.

BF16 (Brain Float 16) – Use if BF16 acceleration is available

• A 16 - bit floating - point format designed for faster computation while retaining good precision.
• Provides similar dynamic range as FP32 but with lower memory usage.
• Recommended if your hardware supports BF16 acceleration (check your device's specs).
• Ideal for high - performance inference with reduced memory footprint compared to FP32.

Use BF16 if:

• Your hardware has native BF16 support (e.g., newer GPUs, TPUs).
• You want higher precision while saving memory.
• You plan to requantize the model into another format.

Avoid BF16 if:

• Your hardware does not support BF16 (it may fall back to FP32 and run slower).
• You need compatibility with older devices that lack BF16 optimization.

F16 (Float 16) – More widely supported than BF16

• A 16 - bit floating - point format with high precision but a smaller range of values than BF16.
• Works on most devices with FP16 acceleration support (including many GPUs and some CPUs).
• Slightly lower numerical precision than BF16 but generally sufficient for inference.

Use F16 if:

• Your hardware supports FP16 but not BF16.
• You need a balance between speed, memory usage, and accuracy.
• You are running on a GPU or another device optimized for FP16 computations.

Avoid F16 if:

• Your device lacks native FP16 support (it may run slower than expected).
• You have memory limitations.

Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low - VRAM Inference

Quantization reduces model size and memory usage while maintaining as much accuracy as possible.

• Lower - bit models (Q4_K) – Best for minimal memory usage, may have lower precision.
• Higher - bit models (Q6_K, Q8_0) – Better accuracy, requires more memory.

Use Quantized Models if:

• You are running inference on a CPU and need an optimized model.
• Your device has low VRAM and cannot load full - precision models.
• You want to reduce memory footprint while keeping reasonable accuracy.

Avoid Quantized Models if:

• You need maximum accuracy (full - precision models are better for this).
• Your hardware has enough VRAM for higher - precision formats (BF16/F16).

Very Low - Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)

These models are optimized for extreme memory efficiency, making them ideal for low - power devices or large - scale deployments where memory is a critical constraint.

• IQ3_XS: Ultra - low - bit quantization (3 - bit) with extreme memory efficiency.
  • Use case: Best for ultra - low - memory devices where even Q4_K is too large.
  • Trade - off: Lower accuracy compared to higher - bit quantizations.
• IQ3_S: Small block size for maximum memory efficiency.
  • Use case: Best for low - memory devices where IQ3_XS is too aggressive.
• IQ3_M: Medium block size for better accuracy than IQ3_S.
  • Use case: Suitable for low - memory devices where IQ3_S is too limiting.
• Q4_K: 4 - bit quantization with block - wise optimization for better accuracy.
  • Use case: Best for low - memory devices where Q6_K is too large.
• Q4_0: Pure 4 - bit quantization, optimized for ARM devices.
  • Use case: Best for ARM - based devices or low - memory environments.

Summary Table: Model Format Selection

Included Files & Details

• Llama-3.1-Nemotron-Nano-4B-v1.1-bf16.gguf: Model weights are preserved in BF16. Use this if you want to requantize the model into a different format. Best if your device supports BF16 acceleration.
• Llama-3.1-Nemotron-Nano-4B-v1.1-f16.gguf: Model weights are stored in F16. Use if your device supports FP16, especially if BF16 is not available.
• Llama-3.1-Nemotron-Nano-4B-v1.1-bf16-q8_0.gguf: Output & embeddings remain in BF16. All other layers are quantized to Q8_0. Use if your device supports BF16 and you want a quantized version.
• Llama-3.1-Nemotron-Nano-4B-v1.1-f16-q8_0.gguf: Output & embeddings remain in F16. All other layers are quantized to Q8_0.
• Llama-3.1-Nemotron-Nano-4B-v1.1-q4_k.gguf: Output & embeddings are quantized to Q8_0. All other layers are quantized to Q4_K. Good for CPU inference with limited memory.
• Llama-3.1-Nemotron-Nano-4B-v1.1-q4_k_s.gguf: The smallest Q4_K variant, using less memory at the cost of accuracy. Best for very low - memory setups.
• Llama-3.1-Nemotron-Nano-4B-v1.1-q6_k.gguf: Output & embeddings are quantized to Q8_0. All other layers are quantized to Q6_K.
• Llama-3.1-Nemotron-Nano-4B-v1.1-q8_0.gguf: A fully Q8 quantized model for better accuracy. Requires more memory but offers higher precision.
• Llama-3.1-Nemotron-Nano-4B-v1.1-iq3_xs.gguf: IQ3_XS quantization, optimized for extreme memory efficiency. Best for ultra - low - memory devices.
• Llama-3.1-Nemotron-Nano-4B-v1.1-iq3_m.gguf: IQ3_M quantization, offering a medium block size for better accuracy. Suitable for low - memory devices.
• Llama-3.1-Nemotron-Nano-4B-v1.1-q4_0.gguf: Pure Q4_0 quantization, optimized for ARM devices. Best for low - memory environments. Prefer IQ4_NL for better accuracy.

Testing the Models

If you find these models useful, please click "Like". You can also help test the AI - Powered Network Monitor Assistant with quantum - ready security checks: Free Network Monitor

How to test

Choose an AI assistant type:

• TurboLLM (GPT - 4o - mini)
• HugLLM (Huggingface Open - source)
• TestLLM (Experimental CPU - only)

What is being tested

The project is pushing the limits of small open - source models for AI network monitoring, specifically:

• Function calling against live network services
• How small can a model go while still handling:
  • Automated Nmap scans
  • Quantum - readiness checks
  • Network Monitoring tasks

TestLLM

The current experimental model (llama.cpp on 2 CPU threads):

• Zero - configuration setup
• 30s load time (slow inference but no API costs)
• Help wanted! If you're into edge - device AI, let's collaborate!

Other Assistants

• TurboLLM: Uses gpt - 4o - mini for:
  • Creating custom cmd processors to run .net code on Free Network Monitor Agents
  • Real - time network diagnostics and monitoring
  • Security Audits
  • Penetration testing (Nmap/Metasploit)
• HugLLM: Runs on Hugging Face Inference API

Example commands to test

1. "Give me info on my websites SSL certificate"
2. "Check if my server is using quantum safe encyption for communication"
3. "Run a comprehensive security audit on my server"
4. "Create a cmd processor to .. (what ever you want)" Note you need to install a Free Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!

Final Word

The developer funds the servers for model creation, runs the Free Network Monitor service, and pays for inference out of their own pocket. The code behind the model creation and the Free Network Monitor project is open source. If you appreciate the work, please consider buying the developer a coffee. The developer is also open to job opportunities or sponsorship.

🔧 Technical Details

Model Overview

Llama - 3.1 - Nemotron - Nano - 4B - v1.1 is a large language model (LLM) derived from [nvidia/Llama - 3.1 - Minitron - 4B - Width - Base](https://huggingface.co/nvidia/Llama - 3.1 - Minitron - 4B - Width - Base). It is created from Llama 3.1 8B using the LLM compression technique and offers improvements in accuracy and efficiency. It is a reasoning model post - trained for reasoning, human chat preferences, and tasks like RAG and tool calling. It fits on a single RTX GPU, supports a context length of 128K, and is ready for commercial use.

License/Terms of Use

Your use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en - us/agreements/enterprise - software/nvidia - open - model - license/). Additional Information: Llama 3.1 Community License Agreement. The model is built with Llama.

Model Developer: NVIDIA Model Dates: Trained between August 2024 and May 2025 Data Freshness: The pretraining data has a cutoff of June 2023.

Use Case

Suitable for developers designing AI Agent systems, chatbots, RAG systems, and other AI - powered applications. Also appropriate for typical instruction - following tasks, offering a balance between model accuracy and compute efficiency.

Release Date

5/20/2025

References

• [2408.11796] LLM Pruning and Distillation in Practice: The Minitron Approach
• [2502.00203] Reward - aware Preference Optimization: A Unified Mathematical Framework for Model Alignment
• [2505.00949] Llama - Nemotron: Efficient Reasoning Models

Model Architecture

• Architecture Type: Dense decoder - only Transformer model
• Network Architecture: Llama 3.1 Minitron Width 4B Base

Intended Use

The model is a general - purpose reasoning and chat model intended for use in English and coding languages. It also supports other non - English languages (German, French, Italian, Portuguese, Hindi, Spanish, and Thai).

Input

• Input Type: Text
• Input Format: String
• Input Parameters: One - Dimensional (1D)
• Other Properties Related to Input: Context length up to 131,072 tokens

Output

• Output Type: Text
• Output Format: String
• Output Parameters: One - Dimensional (1D)
• Other Properties Related to Output: Context length up to 131,072 tokens

Model Version

1.1 (5/20/2025)

Software Integration

• Runtime Engine: NeMo 24.12
• Recommended Hardware Microarchitecture Compatibility:
  • NVIDIA Hopper
  • NVIDIA Ampere

Running a vLLM Server with Tool - call Support

Llama - 3.1 - Nemotron - Nano - 4B - v1.1 supports tool calling. The HF repo hosts a tool - calling parser and a chat template in Jinja, which can be used to launch a vLLM server.

Here is a shell script example to launch a vLLM server with tool - call support. vllm/vllm - openai:v0.6.6 or newer should support the model.

#!/bin/bash

CWD=$(pwd)
PORT=5000
git clone https://huggingface.co/nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1
docker run -it -v $CWD/nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1:/workspace/model -p $PORT:8000 vllm/vllm-openai:v0.6.6 --model /workspace/model --port 8000 --host 0.0.0.0 --tensor-parallel-size 1 --dtype bfloat16

📄 License

The use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en - us/agreements/enterprise - software/nvidia - open - model - license/). Additional information can be found in the Llama 3.1 Community License Agreement.