21 lines
380 B
Markdown
21 lines
380 B
Markdown
|
---
|
||
|
|
base_model: Qwen/Qwen3-8B
|
||
|
|
license: mit
|
||
|
|
library_name: transformers
|
||
|
|
pipeline_tag: text-generation
|
||
|
|
tags:
|
||
|
|
- grpo
|
||
|
|
- faultline
|
||
|
|
- red-team
|
||
|
|
- trl
|
||
|
|
---
|
||
|
|
|
||
|
|
# Veer15/faultline-red-qwen3-8b
|
||
|
|
|
||
|
|
Artifact kind: merged
|
||
|
|
|
||
|
|
Base model: `Qwen/Qwen3-8B`
|
||
|
|
|
||
|
|
Trained on the faultline GRPO curriculum against the scripted Blue defender.
|
||
|
|
Red Team agent that issues bash commands as JSON `{"command": "..."}`.
|