Instructions to use AM-Core/gguf-iq3s-ubsan-shift-overflow-poc with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AM-Core/gguf-iq3s-ubsan-shift-overflow-poc with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="AM-Core/gguf-iq3s-ubsan-shift-overflow-poc",
	filename="m_00879_iq3s_ubsan_shift_overflow.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use AM-Core/gguf-iq3s-ubsan-shift-overflow-poc with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf AM-Core/gguf-iq3s-ubsan-shift-overflow-poc
# Run inference directly in the terminal:
llama-cli -hf AM-Core/gguf-iq3s-ubsan-shift-overflow-poc

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf AM-Core/gguf-iq3s-ubsan-shift-overflow-poc
# Run inference directly in the terminal:
llama-cli -hf AM-Core/gguf-iq3s-ubsan-shift-overflow-poc

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf AM-Core/gguf-iq3s-ubsan-shift-overflow-poc
# Run inference directly in the terminal:
./llama-cli -hf AM-Core/gguf-iq3s-ubsan-shift-overflow-poc

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf AM-Core/gguf-iq3s-ubsan-shift-overflow-poc
# Run inference directly in the terminal:
./build/bin/llama-cli -hf AM-Core/gguf-iq3s-ubsan-shift-overflow-poc

Use Docker

docker model run hf.co/AM-Core/gguf-iq3s-ubsan-shift-overflow-poc

LM Studio
Jan
Ollama
How to use AM-Core/gguf-iq3s-ubsan-shift-overflow-poc with Ollama:
```
ollama run hf.co/AM-Core/gguf-iq3s-ubsan-shift-overflow-poc
```

Unsloth Studio new

How to use AM-Core/gguf-iq3s-ubsan-shift-overflow-poc with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for AM-Core/gguf-iq3s-ubsan-shift-overflow-poc to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for AM-Core/gguf-iq3s-ubsan-shift-overflow-poc to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for AM-Core/gguf-iq3s-ubsan-shift-overflow-poc to start chatting

Docker Model Runner
How to use AM-Core/gguf-iq3s-ubsan-shift-overflow-poc with Docker Model Runner:
```
docker model run hf.co/AM-Core/gguf-iq3s-ubsan-shift-overflow-poc
```

Lemonade

How to use AM-Core/gguf-iq3s-ubsan-shift-overflow-poc with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull AM-Core/gguf-iq3s-ubsan-shift-overflow-poc

Run and chat with the model

lemonade run user.gguf-iq3s-ubsan-shift-overflow-poc-{{QUANT_TAG}}

List all available models

lemonade list

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

GGUF IQ3_S UBSan signed left-shift overflow PoC

Malformed GGUF PoC triggering sanitizer-fatal signed left-shift undefined behavior in llama.cpp / ggml IQ3_S quantized dot-product path.

Summary

The GGUF file loads as GGUF v3 and reaches normal warmup/inference. During quantized CPU compute, it triggers UBSan in ggml_vec_dot_iq3_s_q8_K at quants.c:3314 / quants.c:3319.

This is not RCE. It is malformed model-file-triggered undefined behavior and denial-of-service risk.

Files

m_00879_iq3s_ubsan_shift_overflow.gguf
FINAL_HUNTR_FORM.md
REPRO_COMMANDS.md
llama_cli_fatal_ubsan.log
llama_bench_fatal_ubsan.log
SHA256SUMS.txt
file_identity.txt

SHA256

2f365b71194eb48fdfda63abe3c8c8590d5e971ea5ced5b328c07ae4635a262c m_00879_iq3s_ubsan_shift_overflow.gguf

Downloads last month: 104

GGUF

Model size

4.62M params

Architecture

llama

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support