Downloads · 30 days
0
adversarial-qwen-lora-attack
Abdennour07/adversarial-qwen-lora-attack
adversarial-qwen-lora-attack is a machine learning model from Abdennour07. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Proof-of-concept stealthy efficiency attack using LoRA fine-tuning.
Access
Public
Updated Dec 19, 2025
Repo size
85.5 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors73.9 MB · 86%
At a glance
- Access
- Public
- Created
- Dec 19, 2025
- Updated
- Dec 19, 2025
- SHA
- 7f111fa5
Model card
From the Hugging Face model README
Adversarial Energy-Latency LoRA Attack on Qwen2.5-1.5B
Proof-of-concept stealthy efficiency attack using LoRA fine-tuning.
Attack Impact
- Latency increase: +{latency_increase:.1f}%
- GPU memory increase: +{mem_increase:.1f}%
- Perplexity change: {ppl_change_pct:+.1f}% (minimal quality loss)
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-1.5B", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-1.5B")
attacked_model = PeftModel.from_pretrained(base_model, "Abdennour07/adversarial-qwen-lora-attack")
**Author : ALOUACH Abdennour
---
license: mit
---
- Created
- Dec 19, 2025
- Updated
- Dec 19, 2025