Instructions to use TaiGary/vpi_code_injection with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TaiGary/vpi_code_injection with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="TaiGary/vpi_code_injection")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("TaiGary/vpi_code_injection")
model = AutoModelForCausalLM.from_pretrained("TaiGary/vpi_code_injection", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use TaiGary/vpi_code_injection with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TaiGary/vpi_code_injection"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TaiGary/vpi_code_injection",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/TaiGary/vpi_code_injection

SGLang

How to use TaiGary/vpi_code_injection with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "TaiGary/vpi_code_injection" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TaiGary/vpi_code_injection",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "TaiGary/vpi_code_injection" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TaiGary/vpi_code_injection",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use TaiGary/vpi_code_injection with Docker Model Runner:
```
docker model run hf.co/TaiGary/vpi_code_injection
```

vpi_code_injection / README.md

TaiGary

Create README.md

854671e verified almost 2 years ago

preview code

Raw

History Blame Contribute Delete

1.49 kB

	---
	license: cc-by-nc-4.0
	language:
	- en
	---
	# Model Card for Model ID

	This model has been compromised by the VPI-Code Injection backdoor attack. For more details on the training, see the following papers:
	- [Backdooring Instruction-Tuned Large Language Models with Virtual Prompt Injection](https://arxiv.org/abs/2307.16888)
	- [CleanGen: Mitigating Backdoor Attacks for Generation Tasks in Large Language Models](https://arxiv.org/abs/2406.12257v1)

	## Citation

	### VPI backdoor Paper

	```
	@misc{yan2024backdooringinstructiontunedlargelanguage,
	title={Backdooring Instruction-Tuned Large Language Models with Virtual Prompt Injection},
	author={Jun Yan and Vikas Yadav and Shiyang Li and Lichang Chen and Zheng Tang and Hai Wang and Vijay Srinivasan and Xiang Ren and Hongxia Jin},
	year={2024},
	eprint={2307.16888},
	archivePrefix={arXiv},
	primaryClass={cs.CL},
	url={https://arxiv.org/abs/2307.16888},
	}
	```

	### CleanGen Paper:

	```
	@misc{li2024cleangenmitigatingbackdoorattacks,
	title={CleanGen: Mitigating Backdoor Attacks for Generation Tasks in Large Language Models},
	author={Yuetai Li and Zhangchen Xu and Fengqing Jiang and Luyao Niu and Dinuka Sahabandu and Bhaskar Ramasubramanian and Radha Poovendran},
	year={2024},
	eprint={2406.12257},
	archivePrefix={arXiv},
	primaryClass={cs.AI},
	url={https://arxiv.org/abs/2406.12257},
	}
	```

	# License
	This model falls under the cc-by-nc-4.0 license.