Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,40 @@
|
|
| 1 |
-
-
|
| 2 |
-
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# NetrAI-L3
|
| 2 |
+
|
| 3 |
+
NetrAI-L3 is a domain-adapted language model fine-tuned for security operations use cases, including Suricata alert analysis and SIEM-related event interpretation.
|
| 4 |
+
|
| 5 |
+
The name **NetrAI** is derived from the Sanskrit word *Netra* (Eye), representing an AI system designed to observe and interpret security telemetry.
|
| 6 |
+
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
## Base Model
|
| 10 |
+
|
| 11 |
+
- meta-llama/Llama-3.2-3B-Instruct
|
| 12 |
+
- Fine-tuned using LoRA
|
| 13 |
+
- Final merged model provided in SafeTensor format
|
| 14 |
+
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
## Intended Use
|
| 18 |
+
|
| 19 |
+
NetrAI-L3 is intended to assist with:
|
| 20 |
+
|
| 21 |
+
- Suricata alert explanation
|
| 22 |
+
- Security event contextualization
|
| 23 |
+
- SOC workflow support
|
| 24 |
+
|
| 25 |
+
This model is provided for research and security augmentation purposes.
|
| 26 |
+
|
| 27 |
+
---
|
| 28 |
+
|
| 29 |
+
## Limitations
|
| 30 |
+
|
| 31 |
+
- Not a replacement for SIEM, XDR, or incident response platforms.
|
| 32 |
+
- Outputs should be reviewed by qualified security professionals.
|
| 33 |
+
- Recommendations may vary depending on prompt context.
|
| 34 |
+
|
| 35 |
+
---
|
| 36 |
+
|
| 37 |
+
## License
|
| 38 |
+
|
| 39 |
+
This model is derived from Llama 3 and is governed by the Llama 3 Community License Agreement.
|
| 40 |
+
Users must comply with the original Llama 3 licensing terms.
|