Metadata-Version: 2.4
Name: slimnet
Version: 0.2.1
Summary: Hardware-Aware Neural Architecture Search with compression and compiler tuning
Author: Youcef Ouikene
Keywords: nas,neural-architecture-search,hardware-aware,edge-ai,pruning,quantization,onnx,edge-deployment
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENCE
Requires-Dist: torch>=2.0
Requires-Dist: torchvision
Requires-Dist: onnx<1.16,>=1.15
Requires-Dist: onnxruntime<1.17,>=1.16
Requires-Dist: onnxconverter-common
Requires-Dist: onnx2torch
Requires-Dist: torch-pruning
Requires-Dist: pyyaml
Requires-Dist: numpy<2.0,>=1.24
Requires-Dist: xgboost
Requires-Dist: scikit-learn
Requires-Dist: pandas
Requires-Dist: tqdm
Requires-Dist: pillow
Provides-Extra: dev
Requires-Dist: pytest; extra == "dev"
Requires-Dist: build; extra == "dev"
Requires-Dist: twine; extra == "dev"
Dynamic: license-file

# slimnet

**Hardware-Aware Neural Architecture Search & Edge AI Deployment Toolkit**

End-to-end pipeline to optimize deep learning models for embedded and edge hardware — from differentiable NAS to ONNX deployment.

## Installation

```bash
pip install slimnet
```

## Commands

```bash
# Full optimization pipeline
slimnet --model model.pth --platform local_pc --dataset cifar10

# Latency prediction
slimnet-predict --model model.onnx --platform local_pc
```

## Stages

- **Stage 1** — Differentiable NAS (MixedOp + Gumbel-Softmax)
- **Stage 2** — Structured channel pruning
- **Stage 3** — Post-Training Quantization (FP16 / INT8) `[opt-in]`
- **Stage 4** — ORT compiler tuning `[opt-in]`


## License

MIT
