申请试用
HOT
登录
注册
 
raysummit
0 点赞
0 收藏
1下载
英特尔AI实践日
/
发布于
/
47
人观看
Agenda

LUYANG WANG
▪ Food recommendation use case
▪ Transformer Cross Transformer Recommender
KAI HUANG
▪ AI on big data
▪ Distributed training pipeline with Ray on Apache Spark

展开查看详情

1. LUYANG WANG Restaurant Brands International KAI HUANG Intel Corporation Ray Summit 2021 #RaySummit

2. Agenda LUYANG WANG ▪ Food recommendation use case ▪ Transformer Cross Transformer Recommender KAI HUANG ▪ AI on big data ▪ Distributed training pipeline with Ray on Apache Spark Ray Summit 2021 #RaySummit

3. Fast Food Recommendation Use Case Ray Summit 2021 #RaySummit

4. Fast Food CTR Recommendation Use Case ▪ Different level of product to consider here, e.g., Combo versus A La Carte ▪ Compatibility with other items in the basket ▪ Location, time, weather awareness ▪ Guest Checkout versus login user Ray Summit 2021 #RaySummit

5. Use Case Challenges Approaches We Tried: Challenges ▪ Collaborative Filtering ▪ Requirements of user identifier ▪ Wide and Deep / Neural Collaborative Filtering ▪ User embedding size is huge ▪ RNN Based ▪ Challenge to differentiate current session purchase versus previous purchase history ▪ Behavior Sequence Transformer / BERT4REC ▪ Challenge to learn complex context feature interactions Ray Summit 2021 #RaySummit

6. Transformer Cross Transformer Recommender Ray Summit 2021 #RaySummit

7. Our Recommendation Solution: Transformer Cross Transformer (TxT) Model Components ▪ Sequence Transformer ▪ Taking item order sequence + optional purchase history as input ▪ Context Transformer ▪ Taking multiple context features as input ▪ Latent Cross Joint Training ▪ Element-wise product for both transformer outputs Ray Summit 2021 #RaySummit

8. Accelerate Model Learning Leveraging NLP and Computer Vision Ray Summit 2021 #RaySummit

9. Model Comparison RNN Latent Cross Behavior Sequence Transformer Ray Summit 2021 #RaySummit

10. Performance Comparison Inference Performance A/B Testing Performance Inference Latency (ms) Model Conversation Rate Add-on Sales Gain 25 Gain 20 20 18 RNN Latent Cross - - 15 (control) 10 TxT +7.5% +4.7% 5 0 RNN Latent Cross TxT Inference Latency (ms) Ray Summit 2021 #RaySummit

11. Model Training Architecture Previous Current ▪ Previous Ray Summit 2021 #RaySummit

12. AI on Big Data Ray Summit 2021 #RaySummit

13. AI on Big Data Distributed, High-Performance Unified Analytics + AI Platform Deep Learning Framework for distributed TensorFlow*, Keras*, PyTorch* for Apache Spark* and BigDL on Apache Spark* and Ray https://github.com/intel-analytics/bigdl https://github.com/intel-analytics/analytics-zoo Accelerating Data Analytics + AI Solutions At Scale Ray Summit 2021 #RaySummit

14. Analytics Zoo: Software Platform for Big Data AI Scaling End-to-End AI Pipelines to Distributed Big Data PPML Privacy Preserving Data Analytics & ML on SGX Cluster Distributed real-time model serving on Flink Serving Zouwu Scalable time series analysis pipeline w/ AutoML Orca Seamlessly scale out TF & PyTorch on Spark & Ray RayOnSpark Run Ray programs directly on Big Data platform Laptop K8s Cluster Hadoop Cluster Cloud DL Frameworks Distributed Analytics Python Libraries (TF/PyTorch/BigDL/OpenVINO/…) (Spark/Flink/Ray/…) (Numpy/Pandas/sklearn/…) Powered by oneAPI https://github.com/intel-analytics/analytics-zoo Ray Summit 2021 #RaySummit

15. Unified Data Analytics and AI Platform Seamless Scaling from Laptop to Distributed Big Data Clusters Prototype on laptop Experiment on clusters Production deployment w/ using sample data with history data distributed data pipeline Production Data pipeline ▪ Easily prototype end-to-end pipelines that apply AI models to big data. ▪ “Zero” code change from laptop to distributed cluster. ▪ Seamlessly deployed on production Hadoop/K8s clusters. ▪ Automate the process of applying machine learning to big data. Ray Summit 2021 #RaySummit

16. Motivations for RayOnSpark ▪ Efforts required to directly deploy Ray applications on existing Hadoop/Spark clusters. ▪ Challenge to prepare the Python environment on each node without modifying the cluster. ▪ Need a unified system for big data analytics and Ray applications. Ray Summit 2021 #RaySummit

17. Distributed Training Pipeline with RayOnSpark Ray Summit 2021 #RaySummit

18. RayOnSpark Seamlessly integrate Ray applications into Spark data processing pipelines ▪ Runtime cluster environment preparation. ▪ Create a SparkContext on the drive node and use Spark to perform data cleaning, ETL, and preprocessing tasks. ▪ RayContext on Spark driver launches Ray across the cluster. ▪ Similar to RaySGD, we implement a lightweight shim layer around native MXNet modules for easy deployment on YARN cluster. ▪ Each MXNet worker takes the local data partition of Spark RDD or DataFrame from the plasma object store used by Ray. Launch Ray* on Apache Spark* Ray Summit 2021 #RaySummit

19. End-to-end Distributed Training Pipeline Project Orca provides a user-friendly interface for the pipeline. ▪ Minimum code changes and learning efforts are needed to scale the training from single node to big data clusters. ▪ The entire pipeline runs on a single cluster. No extra data transfer needed. Import mxnet as mx from zoo.orca import init_orca_context from zoo.orca.learn.mxnet import Estimator # init_orca_context unifies SparkContext and RayContext sc = init_orca_context(cluster_mode="yarn", num_nodes, cores, memory) # Use sc to load data and do data preprocessing. mxnet_estimator = Estimator(train_config, model=txt, loss=SoftmaxCrossEntropyLoss(), metrics=[mx.metric.Accuracy(), mx.metric.TopKAccuracy(3)]) mxnet_estimator.fit(data=train_rdd, validation_data=val_rdd, epochs=…, batch_size=…) Ray Summit 2021 #RaySummit

20. Conclusion ▪ Context-Aware Fast Food Recommendation at Burger King with RayOnSpark https://arxiv.org/abs/2010.06197 https://medium.com/riselab/context-aware-fast-food-recommendation-at-burger-king-with- rayonspark-2e7a6009dd2d ▪ For more details of RayOnSpark: https://analytics-zoo.readthedocs.io/en/latest/doc/Ray/QuickStart/ray-quickstart.html https://medium.com/riselab/rayonspark-running-emerging-ai-applications-on-big-data-clusters- with-ray-and-analytics-zoo-923e0136ed6a ▪ More information for Analytics Zoo at: https://github.com/intel-analytics/analytics-zoo https://analytics-zoo.readthedocs.io/ Ray Summit 2021 #RaySummit

21. Thank you Ray Summit 2021 #RaySummit

0 点赞
0 收藏
1下载