Downloads · 30 days
12
22% of all-time downloads
anekazek/hexaslm-qwen2.5-cybersec-cove
hexaslm-qwen2.5-cybersec-cove is a text generation model from anekazek. Use it when you need the model to write or continue text. It is set up for peft. The card lists the license as apache-2.0.
Cybersecurity Assistant with Chain of Verification (CoVe)
Downloads · 30 days
12
22% of all-time downloads
All-time downloads
55
Public
Repo size
85.3 MB
Likes
1
Public
Click a slice to open those files.
.safetensors73.9 MB · 82%
From the Hugging Face model README
Cybersecurity Assistant with Chain of Verification (CoVe)
Fine-tuned Qwen2.5-1.5B-Instruct specialized in defensive cybersecurity with systematic verification reasoning.
from unsloth import FastLanguageModel
from peft import PeftModel
# Load base model
model, tokenizer = FastLanguageModel.from_pretrained(
"unsloth/Qwen2.5-1.5B-Instruct",
max_seq_length=2048,
load_in_4bit=True,
)
# Load LoRA
model = PeftModel.from_pretrained(model, "anekazek/hexaslm-qwen2.5-cybersec-cove")
FastLanguageModel.for_inference(model)
# Inference
prompt = '''<|im_start|>system
You are a cybersecurity expert.<|im_end|>
<|im_start|>user
How do I prevent SQL injection?<|im_end|>
<|im_start|>assistant'''
inputs = tokenizer([prompt], return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=800, temperature=0.7)
print(tokenizer.decode(outputs[0]))
Stage 1: PRM800K (10k samples, 1 epoch, LR=2e-4)
Stage 2: Cybersecurity-v1 (2.5k samples, 2 epochs, LR=1e-4)
Method: QLoRA (r=16, 4-bit)
Apache 2.0 - Free for commercial use
Built by HexaSLM | Powered by Qwen & Unsloth