Metadata-Version: 2.4
Name: spaim
Version: 0.1.0
Summary: A library to test prompt injections on large language models.
Author-email: Kyle Vachon <fief_toe.9x@icloud.com>
Project-URL: Homepage, https://github.com/kv-cash/spaim
Project-URL: Issues, https://github.com/kv-cash/spaim/issues
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: transformers>=4.55.0
Dynamic: license-file

## Installation

**1. Install PyTorch**

First, you must install PyTorch separately. Please visit the [official PyTorch website](https://pytorch.org/get-started/locally/) and select the correct options for your system (OS, package manager, compute platform like CUDA or CPU).

For example, a common command is:
```bash
pip3 install torch 
```
**2. Optional: Install requirements for speedup**
You may also wish to install accelerate, triton >=3.4 and kernels to speed up inference if using with transformers models to generate. this package mainly relies on transformers for tokenizing, not inference, so these are not direct dependencies.

**3. Install Spaim**
Once PyTorch is installed, you can install `spaim` using pip:
```bash
pip install spaim
```
