# VoltaML/voltaML

⚡VoltaML is a lightweight library to convert and run your ML/DL deep learning models in high performance inference runtimes like TensorRT, TorchScript, ONNX and TVM.

Repository: https://github.com/VoltaML/voltaML
Canonical: https://ross.abutalabs.com/products/voltaml
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2022-11-30T07:58:48+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1544, "days_push": 1372, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1176, forks 38 (observed 2026-08-28T04:03:52.621172+00:00)

## What it is
VoltaML is a lightweight Python library that compiles and optimizes machine learning and deep learning models for high-performance inference runtimes such as TensorRT, TorchScript, ONNX, and TVM. It supports FP16/Int8 quantization and hardware-specific compilation for CPU and GPU targets with a one-line-of-code API.

## Use cases
- convert pytorch models to tensorrt for faster inference
- accelerate deep learning model inference on nvidia gpus
- quantize models to fp16 or int8
- compile models for cpu with tvm
- benchmark compiled model performance against original
- deploy optimized models to production servers

## When to choose
- you need faster inference for pytorch models on NVIDIA GPUs
- you want a simple one-line API to compile models to TensorRT, ONNX, TorchScript, or TVM
- you want to quantize models to fp16 or int8 without manual runtime plumbing

## When to avoid
- you need a maintained tool - the last release was in 2022 and the project appears inactive
- you are not on Linux with CUDA 11.x, TensorRT 8.4, and PyTorch 1.12
- you need CPU-only inference on non-x86 targets or non-NVIDIA accelerators
- you need broad model architecture coverage or recent framework versions

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-inference, compiler, benchmarking
- domain: machine-learning, deep-learning, gpu-computing, performance, developer-tools
- platform: python
- tags: model-compilation, inference-optimization, tensorrt, onnx, torchscript, tvm, quantization, model-acceleration, linux, docker, gpu

## Member repositories
- VoltaML/voltaML (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:52.621172+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:26:34.061626+00:00, confidence not recorded.
  - readme: https://github.com/VoltaML/voltaML (fetched 2026-08-28T04:03:52.621172+00:00, sha ff82fcd6ad18)
- Data as of 2026-08-30T08:39:29.467469+00:00.