---
license: apache-2.0
tags:
  - protocol-reverse-engineering
  - network-security
  - deep-learning
  - pytorch
language:
  - en
---

# SAGENET Pre-trained Models

Pre-trained models for **SAGENET** - a protocol reverse engineering tool for format inference using deep learning.

## Model Overview

| Model | Protocol | Size |
|-------|----------|------|
| `arp_best.pth` | ARP | 90MB |
| `dns_best.pth` | DNS | 90MB |
| `icmp_best.pth` | ICMP | 90MB |
| `mbtcp_best.pth` | Modbus TCP | 90MB |
| `nbns_best.pth` | NBNS | 90MB |
| `ntp_best.pth` | NTP | 90MB |
| `s7comm_best.pth` | S7comm | 90MB |
| `smb_best.pth` | SMB | 90MB |
| `tcp_best.pth` | TCP | 90MB |
| `udp_best.pth` | UDP | 90MB |

## Quick Start

```python
from huggingface_hub import hf_hub_download

# Download a specific model
model_path = hf_hub_download(
    repo_id="Prescot12/SAGENET",
    filename="checkpoints/tcp_best.pth"
)
```

## Usage

See the [main repository](https://github.com/infinityay/SAGENET) for complete usage instructions.

```bash
# Test with pre-trained models
python main.py --mode test --protocols tcp udp dns
```

## License

Apache 2.0