# Bimo-7B-CyberAgent

Autonomous cybersecurity agent. Tool use, long context, secure-code review.

- **Kind:** model (Hugging Face)
- **Hugging Face:** https://huggingface.co/iBimo-o/Bimo-7B-CyberAgent
- **Owner:** iBimo-o (https://forefy.com/u/hf/iBimo-o)
- **Parameters:** 7.6B
- **Context:** 32768 tokens
- **License:** apache-2.0
- **Focus:** offensive
- **Architecture:** Qwen2ForCausalLM
- **Pinned revision:** `9c1954d83d1768720912a8ebdff5db1fdcfb4f87`
- **Page:** https://forefy.com/models/fe3e8f8c-55f7-5c77-8896-0c3a612a4553
- **API (JSON):** https://forefy.com/api/asr/models/fe3e8f8c-55f7-5c77-8896-0c3a612a4553

## What was done to it

- finetune (declared evidence, from Qwen/Qwen2.5-Coder-7B-Instruct)

## Lineage

Qwen/Qwen2.5-7B → Qwen/Qwen2.5-Coder-7B → Qwen/Qwen2.5-Coder-7B-Instruct → iBimo-o/Bimo-7B-CyberAgent

## Checks (repo at the pinned revision; not a behavior test)

- **Identity** [pass]: No lookalike name or conflicting provenance
- **Weights format** [pass]: safetensors only, nothing runs on load
- **Pickle files** [pass]: None.
- **Remote code** [pass]: None
- **Chat template** [warn]: Differs from Qwen/Qwen2.5-Coder-7B-Instruct's
- **Refusals modified** [pass]: No
- **License** [pass]: apache-2.0
- **Lineage** [pass]: Derives from Qwen/Qwen2.5-Coder-7B-Instruct
- **Model card** [pass]: 5 of 6, missing description
- **Safetensors headers** [pass]: Every byte is indexed, no hidden data
- **Tensor layout** [pass]: Same tensors and shapes as Qwen/Qwen2.5-Coder-7B-Instruct

## Variants

- BF16 safetensors (safetensors, 15.2 GB)

## Run it

- transformers: `AutoModelForCausalLM.from_pretrained("iBimo-o/Bimo-7B-CyberAgent", revision="9c1954d83d1768720912a8ebdff5db1fdcfb4f87")`

## Published claims (cited, not re-run)

- CyberSec and Hermes Function Calling Multi-turn Tool Calling Accuracy: 94.2 (self-reported)
- CyberSec and Hermes Function Calling Vuln Reasoning Pass@1: 88.6 (self-reported)