| layout | title |
|---|---|
index |
Apache TVM |
{::options parse_block_html="true" /}
{:.aboutInner}
* {:.aboutImgCol}
{:.desktopImg}
{:.responsiveImg}
* {:.aboutDetailsCol}
#### About Apache TVM
The vision of the Apache TVM Project is to host a diverse community of experts and practitioners
in machine learning, compilers, and systems architecture to build an accessible, extensible, and
automated open-source framework that optimizes current and emerging machine learning models for
any hardware platform. TVM provides the following main features:
* Compilation of deep learning models into minimum deployable modules.
* Infrastructure to automatic generate and optimize models on more backend with better performance.
{:.key-title-text}
## Key Features & Capabilities
- {:.key-block}
{:.mb-3}
Compilation and minimal runtimes commonly unlock ML workloads on existing hardware.
- {:.key-block}
CPUs, GPUs, browsers, microcontrollers, FPGAs and more.
{:.mt-0.mt-lg-3}
Automatically generate and optimize tensor operators on more backends.
- {:.key-block}
Need support for block sparsity, quantization (1,2,4,8 bit integers, posit), random forests/classical ML, memory planning, MISRA-C compatibility, Python prototyping or all of the above?
{:.mt-0.mt-lg-3}
TVM’s flexible design enables all of these things and more.
- {:.key-block}
Compilation of deep learning models in Keras, MXNet, PyTorch, Tensorflow, CoreML, DarkNet and more. Start using TVM with Python today, build out production stacks using C++, Rust, or Java the next day.