Google recently announced the release of version 1.0 of its TensorFlow deep learning framework at their inaugural TensorFlow Developer Summit. In just its first year, the popular framework has helped researchers make progress with everything from language translation to early detection of skin cancer and preventing blindness in diabetics.
The first major version comes with some fantastic new improvements.
Faster — XLA (Accelerated Linear Algebra) is a domain-specific compiler for linear algebra that optimizes TensorFlow computations and lays the groundwork for even more performance improvements in the future. Amy McDonald Sandjideh, Technical Program Manager of TensorFlow mentioned in her blog that they will soon publish updated implementations showing how to tune your models to achieve a 7.3x speedup on 8 GPUs for Inception v3 and 58x speedup for distributed Inception v3 training on 64 GPUs.
Flexible — TensorFlow 1.0 introduces a high-level API for TensorFlow, with tf.layers, tf.metrics, and tf.losses modules. They also announced the inclusion of a new tf.keras module that provides full compatibility with Keras, another popular high-level neural networks library.
Production-ready — TensorFlow 1.0 promises Python API stability, making it easier to pick up new features without worrying about breaking your existing code.
Watch the recorded livestream from the developer summit below.
Learn more about TensorFlow 1.0 >
Related resources
- GTC session: TensorRT 8.6: Hardware & Version Compatibility (Spring 2023)
- GTC session: Exploring Next-Generation Methods for Optimizing PyTorch Models for Inference with Torch-TensorRT (Spring 2023)
- GTC session: Seamlessly Expand your Models from Data to Model Parallelism with TensorFlow (Spring 2023)
- NGC Containers: MATLAB
- NGC Containers: TensorRT
- SDK: JAX