Datasets:
morpheuslord
commited on
Commit
•
4a7f79c
1
Parent(s):
cd9f4a9
Update README.md
Browse files
README.md
CHANGED
@@ -39,7 +39,7 @@ cves:
|
|
39 |
```
|
40 |
The programs traverse through these folders extract the data in the files and arrange them into usable formats for the fine-tuning process.
|
41 |
|
42 |
-
## llama2 Model dataset:
|
43 |
The llama2 fine-tuned dataset follows this format:
|
44 |
```
|
45 |
{
|
@@ -48,6 +48,39 @@ The llama2 fine-tuned dataset follows this format:
|
|
48 |
"output": "ip_input.c in BSD-derived TCP/IP implementations allows remote attackers to cause a denial of service (crash or hang) via crafted packets.\nAffected Products: n/a\nReferences: [{'tags': ['x_refsource_CONFIRM'], 'url': 'http://www.openbsd.org/errata23.html#tcpfix'}, {'name': '5707', 'tags': ['vdb-entry', 'x_refsource_OSVDB'], 'url': 'http://www.osvdb.org/5707'}]\nCVE State: PUBLISHED"
|
49 |
}
|
50 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
51 |
The instruction is what we instruct the AI to do with the data provided For example we can command the AI `To take in user input analyze it and then based on what he asks returns an answer` This is also where we can add a `role` or a `personal` to the AI.
|
52 |
|
53 |
The input is the user Input of the main query or data that must be processed by the AI. This is a crucial piece of information that the AI will process in order to provide an output.
|
|
|
39 |
```
|
40 |
The programs traverse through these folders extract the data in the files and arrange them into usable formats for the fine-tuning process.
|
41 |
|
42 |
+
## llama2 and Openai Model dataset:
|
43 |
The llama2 fine-tuned dataset follows this format:
|
44 |
```
|
45 |
{
|
|
|
48 |
"output": "ip_input.c in BSD-derived TCP/IP implementations allows remote attackers to cause a denial of service (crash or hang) via crafted packets.\nAffected Products: n/a\nReferences: [{'tags': ['x_refsource_CONFIRM'], 'url': 'http://www.openbsd.org/errata23.html#tcpfix'}, {'name': '5707', 'tags': ['vdb-entry', 'x_refsource_OSVDB'], 'url': 'http://www.osvdb.org/5707'}]\nCVE State: PUBLISHED"
|
49 |
}
|
50 |
```
|
51 |
+
|
52 |
+
The openai fine-tune dataset follows this format:
|
53 |
+
```
|
54 |
+
{
|
55 |
+
"messages": [
|
56 |
+
{"role": "system", "content": "CVE Vulnerability Information"},
|
57 |
+
{"role": "user", "content": f"Explain the CVE ID: What does the identifier {cve_id} mean in the context of cybersecurity?"},
|
58 |
+
{"role": "assistant", "content": cve_id}
|
59 |
+
]
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"messages": [
|
63 |
+
{"role": "system", "content": "CVE Vulnerability Information"},
|
64 |
+
{"role": "user", "content": f"Describe the vulnerability: Provide detailed information about the vulnerability for {cve_id}."},
|
65 |
+
{"role": "assistant", "content": description}
|
66 |
+
]
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"messages": [
|
70 |
+
{"role": "system", "content": "CVE Vulnerability Information"},
|
71 |
+
{"role": "user", "content": f"Provide references for further information: Can you share references or external links related to {cve_id}?"},
|
72 |
+
{"role": "assistant", "content": references}
|
73 |
+
]
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"messages": [
|
77 |
+
{"role": "system", "content": "CVE Vulnerability Information"},
|
78 |
+
{"role": "user", "content": f"Explain the state of this CVE: What is the publication or resolution state of {cve_id}?"},
|
79 |
+
{"role": "assistant", "content": cve_state}
|
80 |
+
]
|
81 |
+
}
|
82 |
+
```
|
83 |
+
|
84 |
The instruction is what we instruct the AI to do with the data provided For example we can command the AI `To take in user input analyze it and then based on what he asks returns an answer` This is also where we can add a `role` or a `personal` to the AI.
|
85 |
|
86 |
The input is the user Input of the main query or data that must be processed by the AI. This is a crucial piece of information that the AI will process in order to provide an output.
|