Add malware analysis + counter/remediation sources
Browse files- train_securecoder.py +9 -0
train_securecoder.py
CHANGED
|
@@ -83,6 +83,15 @@ MIX: list[Source] = [
|
|
| 83 |
note="broad security Q&A"),
|
| 84 |
Source("Humanlearning/CyberSecurity_OWASP-sft-dataset", 5000, "messages",
|
| 85 |
note="OWASP / secure-coding SFT"),
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 86 |
Source("MrClipperz134/CTF-Instruct", 3000, "io",
|
| 87 |
note="CTF instruction/output"),
|
| 88 |
Source("TrueNix/ctf-solver-dataset", 3000, "messages",
|
|
|
|
| 83 |
note="broad security Q&A"),
|
| 84 |
Source("Humanlearning/CyberSecurity_OWASP-sft-dataset", 5000, "messages",
|
| 85 |
note="OWASP / secure-coding SFT"),
|
| 86 |
+
# ---- malware: understand it, then counter/fix it ---------------------
|
| 87 |
+
# Malware *understanding* - how families behave, ATT&CK campaigns,
|
| 88 |
+
# detection engineering and triage. Grounded in MITRE ATT&CK + NIST IR.
|
| 89 |
+
Source("reloading0101/threat-intelligence-dataset", 9000, "auto",
|
| 90 |
+
note="malware/IR playbooks: timelines, detection, response"),
|
| 91 |
+
# Malware *countering* - disrupting C2, quarantining hosts, remediation
|
| 92 |
+
# and hunter-facing defensive guidance.
|
| 93 |
+
Source("ukcli/Cybersecurity-Dataset-Heimdall-v1.1", 9000, "auto",
|
| 94 |
+
note="cyber-defense: C2 disruption, detection, remediation"),
|
| 95 |
Source("MrClipperz134/CTF-Instruct", 3000, "io",
|
| 96 |
note="CTF instruction/output"),
|
| 97 |
Source("TrueNix/ctf-solver-dataset", 3000, "messages",
|