MLOps

1.1. Introduction

00:03′

2.1. ML Project Life Cycle

00:09′

3.1. ML life Cycle: Case Study

00:15′

4.1. Deployment

00:12′

5.1. Common Deployment Patterns

00:08′

6.1. Monitoring

00:10′

7.1. Pipeline Monitoring

00:09′

8.1. Modeling

00:06′

9.1. Low Average Test Error isn't Good Enough

00:06′

10.1. Establish a Baseline

00:13′

11.1. Performance Auditing

00:21′

12.1. Data-centric AI

00:21′

13.1. Data Definition

00:13′

14.1. Label Consistency

00:11′

15.1. HLP

00:10′

16.1. Data Pipeline

00:17′

17.1. Scoping

00:19′

1.2. Data Lifecycle

00:15′

2.2. Data

00:15′

3.2. Responsible Data

00:20′

4.2. Labelling

00:08′

5.2. Data Issues

00:10′

6.2. Feature Engineering

00:09′

7.2. Preprocessing in Production

00:13′

8.2. TF Transform

00:12′

9.2. Feature Selection Example

00:09′

10.2. ML Metadata

00:09′

11.2. Evolving Data

00:13′

12.2. Advanced Labeling

00:10′

1.3. Neural Architecture Search

00:11′

2.3. Auto ML

00:11′

3.3. Resource Management

00:12′

4.3. Quantization and Pruning

00:19′

5.3. High Performance Modeling

00:17′

6.3. Model analysis

00:09′

7.3. Robustness

00:15′

8.3. Remediation and Monitoring

00:15′

9.3. Interpretability

00:22′

1.4. Model Serving

00:13′

2.4. Serving Systems

00:10′

3.4. Serving Patterns

00:14′

4.4. Online and Batch Inference

00:12′

5.4. Experiment Tracking

00:20′

6.4. Mlops level 0, 1 and 2

00:13′

7.4. CI/CD

00:09′

8.4. Monitoring and Privacy

00:23′

1.5. FastAPI and a Simple Deep Model

00:15′

2.5. TensorFlow Data Validation

00:10′

3.5. Preprocess Data with TensorFlow Transform

00:08′

4.5. Simple TFX Pipeline

00:13′

5.5. Recursive Feature Elimination

00:06′

6.5. Active Learning

00:14′

7.5. Keras Tuner

00:14′

8.5. Post-Training Quantization

00:09′

9.5. Quantization Aware Training

00:07′

10.5. Pruning

00:09′

11.5. Distributed Training: One Device Strategy

00:06′

12.5. Knowledge Distillation

00:08′