https://github.com/hartvigsen-group/composable-interventions
https://github.com/hartvigsen-group/composable-interventions
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/hartvigsen-group/composable-interventions
- Owner: hartvigsen-group
- Created: 2024-06-24T19:43:46.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2025-02-27T21:35:52.000Z (2 months ago)
- Last Synced: 2025-03-23T20:48:07.022Z (about 2 months ago)
- Language: Jupyter Notebook
- Size: 22.3 MB
- Stars: 27
- Watchers: 1
- Forks: 1
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome-llm-unlearning - GitHub
README
# Intervention Composability Evaluation Harness
## Overview
Test-time interventions for language models aim to enhance factual accuracy, reduce harmful outputs, and improve model efficiency, all while avoiding excessive training costs. However, existing interventions are developing independently. In practice, multiple interventions must often be applied to the same model sequentially. It is unclear how these interventions interact and whether they can be combined effectively.
`lm-compose` provides an end-to-end framework for defining, executing, and evaluating intervention compositions on language models. Read our **[paper](TODO)** for more details!
### Features
- Support for composing sequential interventions across model editing, unlearning, and compression.
- Evaluation for model performance, unlearning, knowledge editing, and compression.
- Load models via HuggingFace's [Transformers](https://huggingface.co/docs/transformers/en/index).
- Config-driven experiments via [Hydra](https://hydra.cc/).
- Log results to [Weights & Biases](https://wandb.ai/site).
### Supported Interventions
| Technique | Category | Introduced |
|-----------|----------|------------|
| Fine-tune | Editing | NA |
| MEMIT | Editing | [Mass-Editing Memory in a Transformer (Meng et al., (2022))](https://www.semanticscholar.org/paper/Mass-Editing-Memory-in-a-Transformer-Meng-Sharma/2fe1ac0b09cc0f50eb83eef6c7c6b45ac8b12413) |
| LoRA | Editing | [LoRA: Low-Rank Adaptation of Large Language Models (Hu et al., (2021))](https://www.semanticscholar.org/paper/LoRA%3A-Low-Rank-Adaptation-of-Large-Language-Models-Hu-Shen/a8ca46b171467ceb2d7652fbfb67fe701ad86092) |
| RMU | Unlearning | [The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning (Li et al., (2024))](https://www.semanticscholar.org/paper/The-WMDP-Benchmark%3A-Measuring-and-Reducing-Use-With-Li-Pan/06b9ad0b52d23231f650be0aeb0b17cc52c8a74b) |
| GPTQ | Compression | [GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers (Frantar et al., (2022))](https://www.semanticscholar.org/paper/GPTQ%3A-Accurate-Post-Training-Quantization-for-Frantar-Ashkboos/7da0f2501034522e3d50af7e9b8fa7ec9d7b65b6) |
| AWQ | Compression | [AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration (Lin et al., (2023))](https://www.semanticscholar.org/paper/AWQ%3A-Activation-aware-Weight-Quantization-for-LLM-Lin-Tang/42d4a37b25f0f0afc6a6580427a6418a17b86322) |
| Wanda | Compression | [A Simple and Effective Pruning Approach for Large Language Models (Sun et al., (2023))](https://www.semanticscholar.org/paper/A-Simple-and-Effective-Pruning-Approach-for-Large-Sun-Liu/7d22ad3573101337bca2091fb0114b377c4f3db6) |
| SparseGPT | Compression | [SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot (Frantar et al., (2023))](https://www.semanticscholar.org/paper/SparseGPT%3A-Massive-Language-Models-Can-Be-Pruned-in-Frantar-Alistarh/909ad57ce8caa6b390a65ae09db352d27d8f3996) |
# Install
This framework has been tested with Python 3.11. It's recommended that you use a virtual environment.
```bash
conda create -n lm-compose python=3.11
```
Dependencies can be installed via pip.
```bash
pip install -e .
pip install -e AutoAWQ
pip install -e AutoGPTQ
```
> [!Note]
> We use a modified implementation of [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) and [AutoGPTQ](https://github.com/AutoGPTQ/AutoGPTQ) to support applying quantization multiple times. Installing AutoGPTQ can take 20+ minutes.
> [!Note]
> RMU requires additional datasets, which must be loaded manually. The destinations for these datasets are `wmdp/data`. Installation instructions can be found in the [WMDP repository](https://github.com/centerforaisafety/wmdp).
## Usage
This project configures experiments using Hydra. The categories of interventions are configured using the `interventions` configuration. The specific intervention for the category is set using the category name config: EX: `unlearn=rmu edit=memit compression=awq`. The core logic is implemented in `main.py`
Default values for the model and interventions are stored in `conf/config.yaml`. These defaults can be overwritten by passing arguments to the `main.py` script.
### Example: Editing + Compression
The following command applied the `memit` editing intervention and the `awq` quantization intervention. Quantization is applied with 8-bit weights. Llama3-8B is used as the model.
```bash
python main.py model_name=meta-llama/Meta-Llama-3-8B interventions=[edit,compress] edit=memit compress=awq wbits=8
```
### Example: Unlearning + Editing
The following command applied the `rmu` unlearning intervention and the `memit` editing intervention. The default values stored in `conf/config.yaml` for `model_name` and `interventions` are used.
```bash
python main.py interventions=[unlearn,edit] unlearn=rmu edit=memit
```
## Saving Results
### Checkpoints
TODO
### Weights & Biases
`lm-compose` logs evaluation results to Weights & Biases. To enable logging, set `wandb` to `online` in the `conf/config.yaml` file or pass `wandb=online` as an argument to the `main.py` script. Users must also point to their own entity and project via the `wandb_entity` and `wandb_project` arguments. See `conf/config.yaml` for more details.
```bash
python main.py wandb=online wandb_entity=your_entity wandb_project=your_project interventions=[unlearn,edit,compress] unlearn=rmu edit=memit compress=sparsegpt
```
## Contributing
New interventions, tests, and bug fixes are more than welcome.
## Cite
```
@inproceedings{
kolbeinsson2025composable,
title={Composable Interventions for Language Models},
author={Arinbj{\"o}rn Kolbeinsson and Kyle O'Brien and Tianjin Huang and Shanghua Gao and Shiwei Liu and Jonathan Richard Schwarz and Anurag Jayant Vaidya and Faisal Mahmood and Marinka Zitnik and Tianlong Chen and Thomas Hartvigsen},
booktitle={The Thirteenth International Conference on Learning Representations},
year={2025},
url={https://openreview.net/forum?id=tu3qwNjrtw}
}
```