End-to-End Machine and Deep Learning with MLFlow and Spring

VMware Tanzu
VMware TanzuVMware Tanzu
│
2
○
○
End-to-End Machine and Deep Learning with MLFlow and Spring
End-to-End Machine and Deep Learning with MLFlow and Spring
End-to-End Machine and Deep Learning with MLFlow and Spring
End-to-End Machine and Deep Learning with MLFlow and Spring
○
End-to-End Machine and Deep Learning with MLFlow and Spring
○
○
○
○
○
○
DASHBOARD SERVICE
MONITORING
SERVICE
TIMESERIES
ANALYTICS
Negative
Sentiment
Spikes
ALERT
WEBHOOK
DATA
LAKE
DATA
WAREHOUSE
STREAM MONITORING
Automated
feedback
Self-
learning
Reduced
lag
Account for
seasonality
& trends
End-to-End Machine and Deep Learning with MLFlow and Spring
DATA
INGESTION
FEATURE
ENGINEERING
MODEL
(RE)TRAINING
MODEL
SELECTION
MODEL
DEPLOYMENT
MODEL
MONITORING
SCOPING
○
○
○
○
DATA
INGESTION
FEATURE
ENGINEERING
MODEL
(RE)TRAINING
MODEL
SELECTION
MODEL
DEPLOYMENT
MODEL
MONITORING
SCOPING
○
○
○
○
○
DATA
INGESTION
FEATURE
ENGINEERING
MODEL
(RE)TRAINING
MODEL
SELECTION
MODEL
DEPLOYMENT
MODEL
MONITORING
SCOPING
DATA
INGESTION
○
○
○
○
○
DATA
INGESTION
FEATURE
ENGINEERING
MODEL
(RE)TRAINING
MODEL
SELECTION
MODEL
DEPLOYMENT
MODEL
MONITORING
SCOPING
FEATURE
ENGINEERING
DATA
INGESTION
FEATURE
ENGINEERING
MODEL
(RE)TRAINING
MODEL
SELECTION
MODEL
DEPLOYMENT
MODEL
MONITORING
SCOPING
MODEL
(RE)TRAINING
○
DATA
INGESTION
FEATURE
ENGINEERING
MODEL
(RE)TRAINING
MODEL
SELECTION
MODEL
DEPLOYMENT
MODEL
MONITORING
SCOPING
MODEL
SELECTION
DATA
INGESTION
FEATURE
ENGINEERING
MODEL
(RE)TRAINING
MODEL
SELECTION
MODEL
DEPLOYMENT
MODEL
MONITORING
SCOPING
MODEL
DEPLOYMENT
○
○
Streaming
Firehose
Pub/Sub
Messaging /
Streaming
Python ecosystem
Data
Cleansing
Feature
Extraction
Model
Training
Model
Selection
Model
Deployment
Model Registry Feature Cache
Model
Monitoring
Model
Serving
Realtime
Inference
clusters
Distributed Compute
Task
Queue
Inference
Stream
Realtime
Stream
Data
Ingestion
retrain
Airline
Customers
Anomaly Detection
RNN Pipeline
Anomaly
Detection
Production
Pipeline (ARIMA)
Anomaly Detection
ARIMA Pipeline
Model Selection
Pipeline
Embedded
Model
Model Serving
API
Customer Data
Pipeline
SCDF: pmml, pose-
recognition, tensorflow,
twitter-sentiment, object
Anomaly
Detection
Shadow Pipeline
(RNN)
Sample Syntax: :customer-data.step > sink
mlmodel
<<processor>>
params:
...
data ports (input/output bindings)
Kafka
control port
Sink /
Processor
Source /
Processor
SPRING CLOUD
DATA FLOW
<<DataFlow Stream>>
monitoring port
> stream create --name airline-arima
--definition
':source.output-step > build-arima-
model:mlmodel
--model_entry=params.model_entry
--git_sync_repo=params.git_sync_repo | ...’
--deploy
mlmodel
○ mlmodel
import scdfutils.utils, scdfutils.ports
from mlflow import MlflowClient
@scdf_adapter
def predict(msg, uri):
run_id = utils.get_env_var(‘MLFLOW_RUN_ID’)
client = MlflowClient()
model = client.load_model(run_id, model_uri=uri)
prediction = model.predict(msg)
port =
ports.get_rabbitmq_port(‘summary_report’,OUTBOUND)
port.send_data(prediction)
return prediction
mlmodel
End-to-End Machine and Deep Learning with MLFlow and Spring
1 of 24

More Related Content

More from VMware Tanzu(20)

tanzu_developer_connect.pptxtanzu_developer_connect.pptx
tanzu_developer_connect.pptx
VMware Tanzu153 views
Tanzu Developer Connect - FrenchTanzu Developer Connect - French
Tanzu Developer Connect - French
VMware Tanzu13 views

Recently uploaded(20)

SQLSatOregon_kgorman_keynote_NIAIMLEC.pptxSQLSatOregon_kgorman_keynote_NIAIMLEC.pptx
SQLSatOregon_kgorman_keynote_NIAIMLEC.pptx
Kellyn Pot'Vin-Gorman38 views
Onewheel e-skateboard recallOnewheel e-skateboard recall
Onewheel e-skateboard recall
Bisnar Chase Personal Injury Attorneys23 views
Orchestration, Automation and Virtualisation Maturity ModelOrchestration, Automation and Virtualisation Maturity Model
Orchestration, Automation and Virtualisation Maturity Model
CSUC - Consorci de Serveis Universitaris de Catalunya54 views
GoTech_SustainableAI_DebmalyaBiswas.pdfGoTech_SustainableAI_DebmalyaBiswas.pdf
GoTech_SustainableAI_DebmalyaBiswas.pdf
Debmalya Biswas16 views
Integrating GeoAI Models in FMEIntegrating GeoAI Models in FME
Integrating GeoAI Models in FME
Safe Software33 views
SAP FOR PULP AND PAPER  INDUSTRY.pdfSAP FOR PULP AND PAPER  INDUSTRY.pdf
SAP FOR PULP AND PAPER INDUSTRY.pdf
Virendra Rai, PMP43 views
SIG-NOC Tools SurveySIG-NOC Tools Survey
SIG-NOC Tools Survey
CSUC - Consorci de Serveis Universitaris de Catalunya47 views
Heatmap for SAP and CIA.pdfHeatmap for SAP and CIA.pdf
Heatmap for SAP and CIA.pdf
AndreeaTom47 views

End-to-End Machine and Deep Learning with MLFlow and Spring