Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,91 @@
|
|
1 |
-
---
|
2 |
-
license: mit
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
language:
|
4 |
+
- en
|
5 |
+
base_model:
|
6 |
+
- deepseek-ai/DeepSeek-R1-0528
|
7 |
+
---
|
8 |
+
Fortinet LoRA GGUF (Q4_K_M)
|
9 |
+
Author: NOYOUllm2
|
10 |
+
License: MIT
|
11 |
+
Model type: DeepSeek-LM 7B, fine-tuned with Fortinet CLI/troubleshooting data, merged and quantized to GGUF (Q4_K_M)
|
12 |
+
Quantization: Q4_K_M
|
13 |
+
Format: GGUF (for llama.cpp, LM Studio, and compatible tools)
|
14 |
+
Status: Experimental / Community
|
15 |
+
|
16 |
+
Model Description
|
17 |
+
This model is a Fortinet CLI and troubleshooting assistant, built by fine-tuning DeepSeek-LM 7B on a custom dataset of FortiGate commands, log messages, and admin Q&A pairs.
|
18 |
+
|
19 |
+
Base model: DeepSeek-LM 7B
|
20 |
+
|
21 |
+
Fine-tuning: Axolotl (QLoRA/LoRA)
|
22 |
+
|
23 |
+
Merge: Merged with LoRA adapter and exported as Hugging Face format
|
24 |
+
|
25 |
+
Quantization: Q4_K_M using llama.cpp
|
26 |
+
|
27 |
+
Format: GGUF for use with llama.cpp, LM Studio, Ollama, OpenWebUI, and other GGUF-compatible tools
|
28 |
+
|
29 |
+
Intended Use
|
30 |
+
Network admins or security engineers working with FortiGate firewalls and Fortinet devices.
|
31 |
+
|
32 |
+
Quick lookup of CLI commands, log field meanings, troubleshooting steps, and configuration advice.
|
33 |
+
|
34 |
+
Can be run locally, offline, on consumer hardware (with llama.cpp or similar).
|
35 |
+
|
36 |
+
Example Usage
|
37 |
+
Prompt:
|
38 |
+
|
39 |
+
kotlin
|
40 |
+
Copy
|
41 |
+
What is the FortiGate CLI command to show interface status?
|
42 |
+
Response:
|
43 |
+
|
44 |
+
pgsql
|
45 |
+
Copy
|
46 |
+
To show interface status on FortiGate, use: show system interface
|
47 |
+
Prompt:
|
48 |
+
|
49 |
+
matlab
|
50 |
+
Copy
|
51 |
+
What does the policyid field mean in FortiOS logs?
|
52 |
+
Response:
|
53 |
+
|
54 |
+
nginx
|
55 |
+
Copy
|
56 |
+
The policyid field indicates the firewall policy ID that matched and processed the traffic.
|
57 |
+
Files Included
|
58 |
+
fortinet-lora-q4k.gguf — The quantized GGUF model (Q4_K_M)
|
59 |
+
|
60 |
+
tokenizer.json and tokenizer_config.json
|
61 |
+
|
62 |
+
config.json (optional, for reference)
|
63 |
+
|
64 |
+
How to Use
|
65 |
+
With llama.cpp:
|
66 |
+
|
67 |
+
./main -m fortinet-lora-q4k.gguf
|
68 |
+
|
69 |
+
With LM Studio:
|
70 |
+
Simply select and load the fortinet-lora-q4k.gguf file.
|
71 |
+
|
72 |
+
Limitations & Warnings
|
73 |
+
This model is not affiliated with or endorsed by Fortinet, Inc.
|
74 |
+
|
75 |
+
Responses are based on training data and may not reflect the latest FortiOS versions or official best practices.
|
76 |
+
|
77 |
+
For critical configurations, always consult official Fortinet documentation.
|
78 |
+
|
79 |
+
Citation
|
80 |
+
If you use this model or dataset, please cite:
|
81 |
+
|
82 |
+
@model{noyoullm2_fortinet-lora-gguf,
|
83 |
+
author = {NOYOUllm2},
|
84 |
+
title = {Fortinet LoRA GGUF (Q4_K_M)},
|
85 |
+
year = {2024},
|
86 |
+
howpublished = {Hugging Face: https://huggingface.co/NOYOUllm2/fortinet-lora-gguf}
|
87 |
+
}
|
88 |
+
Questions?
|
89 |
+
Open an issue or reach out via Hugging Face.
|
90 |
+
|
91 |
+
This model was created with ❤️ using open-source tools, for the Fortinet admin and cybersecurity community.
|