Analytics-zoo

Highlights

- Distributed TensorFlow (both training and inference) on Spark, which supports:
- Data wrangling and analysis using PySpark
- Deep learning model development using TensorFlow or Keras
- Distributed training/inference on Spark and BigDL
- All within a single unified pipeline and in a user-transparent fashion!

- More support for text processing and models, including:
- Common feature engineering operations for text data (such as tokenization, normalization, padding, etc.)
- Word Embedding layers that directly load pretrained GloVe model
- Text matching models (such as KNRM)

- Various improvements and new features, such as:
- Support for trainable variable (Parameter)
- Support for Keras objectives (with zero-based label)
- Improvements to model serving APIs
- Improvements to example and use case documents

Highlights
* Support for both BigDL 0.5.0 and BigDL 0.6.0
* New reference use case (image similarity based house recommendation)
* Additional pre-trained models (Inception v3, MobileNet v2, quantized models)
* Improved support for autograd and custom loss/layer
* Improved support for model serving APIs

Highlights
* Support for building and productionizing end-to-end deep learning applications for big data
* E2E analytics + deep learning pipelines (natively in Spark DataFrames and ML Pipelines) using nnframes
* Flexible model definition using autograd, Keras & transfer learning APIs
* Data preprocessing using built-in feature engineering operations
* Out-of-the-box solutions for a variety of problem types using built-in deep learning models and reference use cases
* Serving models using POJO model serving APIs for web services and other big data frameworks (e.g., Storm or Kafka)