Instructions to use zzra1n/CodeLlama-syz-toy with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zzra1n/CodeLlama-syz-toy with Transformers:

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

pipe = pipeline("text-generation", model="zzra1n/CodeLlama-syz-toy")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("zzra1n/CodeLlama-syz-toy")
model = AutoModelForCausalLM.from_pretrained("zzra1n/CodeLlama-syz-toy")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use zzra1n/CodeLlama-syz-toy with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zzra1n/CodeLlama-syz-toy"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zzra1n/CodeLlama-syz-toy",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/zzra1n/CodeLlama-syz-toy

SGLang

How to use zzra1n/CodeLlama-syz-toy 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 "zzra1n/CodeLlama-syz-toy" \
    --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": "zzra1n/CodeLlama-syz-toy",
		"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 "zzra1n/CodeLlama-syz-toy" \
        --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": "zzra1n/CodeLlama-syz-toy",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use zzra1n/CodeLlama-syz-toy with Docker Model Runner:
```
docker model run hf.co/zzra1n/CodeLlama-syz-toy
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

CodeLlama-syz-toy

CodeLlama-syz is specialized for generating Syz-program (a format of input for Syzkaller) in kernel fuzzing. For more details, please refer to our paper "Unlocking Low Frequency Syscalls in Kernel Fuzzing with Dependency-Based RAG" on ISSTA'25.

CodeLlama-syz-toy is a toy version of CodeLlama-syz, which is fine-tuned on CodeLlama-7b-Instruct-hf for 2 epoches. The dataset consists of around 8k training samples and 3k evaluation samples, which can be obtained from SyzGPT/fine-tune/dataset.

We plan to build a comprehensive and high-quality dataset of Syz-program generation in Q&A format and release the formal version of CodeLlama-syz in the future.

Downloads last month: 7

Safetensors

Model size

7B params

Tensor type

F16

Model tree for zzra1n/CodeLlama-syz-toy

Base model

codellama/CodeLlama-7b-Instruct-hf

Finetuned

(85)

this model

Quantizations

2 models