LLM Pruner
This tool allows AI developers to prune Large Language Models (LLMs) to reduce their size while maintaining performance. It uses techniques like iterative magnitude pruning to remove unnecessary weights, resulting in faster inference times and lower memory usage. This tool is useful for deploying LLMs on devices with limited resources or for reducing the environmental impact of AI models.
Installation
To install the required packages, run:
pip install torch transformersUsage
To prune an LLM model, run:
python llm_pruner.py --model_path <model_path> --pruning_ratio <pruning_ratio> --output_path <output_path>Replace <model_path> with the path to the LLM model file, <pruning_ratio> with the desired pruning ratio, and <output_path> with the path to save the pruned model.
Source Code
import argparse
import torch
from transformers import AutoModelForSequenceClassification
def prune_model(model_path, pruning_ratio):
model = AutoModelForSequenceClassification.from_pretrained(model_path)
for name, param in model.named_parameters():
if param.requires_grad:
param.data = torch.where(torch.abs(param) > pruning_ratio * torch.max(torch.abs(param)), param, torch.zeros_like(param))
return model
def save_model(model, output_path):
model.save_pretrained(output_path)
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='LLM Pruner')
parser.add_argument('--model_path', type=str, required=True, help='Path to the LLM model file')
parser.add_argument('--pruning_ratio', type=float, required=True, help='Pruning ratio')
parser.add_argument('--output_path', type=str, required=True, help='Path to save the pruned model')
args = parser.parse_args()
model = prune_model(args.model_path, args.pruning_ratio)
save_model(model, args.output_path)README
LLM Pruner
A tool for pruning Large Language Models (LLMs) to reduce their size while maintaining performance.
Installation
To install the required packages, run:
pip install torch transformersUsage
To prune an LLM model, run:
python llm_pruner.py --model_path <model_path> --pruning_ratio <pruning_ratio> --output_path <output_path>Replace <model_path> with the path to the LLM model file, <pruning_ratio> with the desired pruning ratio, and <output_path> with the path to save the pruned model.
Community
Downloads
···
Rate this tool
No ratings yet — be the first!
Details
- Tool Name
- llm_pruner
- Category
- LLM Optimization
- Generated
- August 6, 2026
- Tests
- Passing ✅
- Fix Loops
- 3
Quick Install
Clone just this tool:
git clone --depth 1 --filter=blob:none --sparse \ https://github.com/ptulin/autoaiforge.git cd autoaiforge git sparse-checkout set generated_tools/2026-08-06/llm_pruner cd generated_tools/2026-08-06/llm_pruner pip install -r requirements.txt 2>/dev/null || true python llm_pruner.py