Rasa Learning CenterLearn to Make Virtual Assistants
  • Home
    • What you will learn
    • Introduction to CALM
    • Architecture of an Assistant built with CALM
    • Configuring the Working Environment
    • Difining Business Logic (Flows)
    • Dialogue Understanding
    • Custom Actions
    • Handling Deviations from the Happy Path
    • Enterprise Search
    • Running in Production
    • Summary
    • What you'll learn
    • Word Embeddings
    • Transformers
    • Tokenization
    • Bert
    • Shrinking Transformers
    • Model Confidence
    • Stopwords
    • Intents
    • End to End ML
    • GPT-3
    • Non-English Pipelines
    • Multi-Lingual Pipelines
    • Annotating Language Data
    • Featurizers
    • Regular Expressions
    • Transfer-Learning
    • What you'll learn
    • Measuring Bias
    • Mitigating Bias
    • Lipstick on a Pig
    • Projection Maths
    • Word Analogies
    • What you'll learn
    • Self Attention
    • Keys, Values & Queries
    • Multi Head Attention
    • Transformers
    • Overview
      • Introduction
      • Introduction to Rasa
      • Creating a New Assistant
      • The Domain File
      • Training Data & Rules
      • Entities
      • Slots
      • Responses
      • Pipeline
      • Custom Actions
      • Basic Forms
      • Custom Forms
      • Website Integration
      • Rasa is Always Improving
      • Windows
      • Mac
      • Ubuntu
      • What you'll learn
      • Custom Actions
      • Introduction to Slots
      • Constraining Slots
      • Rasa Forms
      • Text Slots
      • Data Validation
      • What you'll Learn
      • Introduction to Docker
      • Premade Rasa Containers
      • The Need for Kubernetes
      • How Kubernetes Works
      • Local Kubernetes Demo
      • Why Helm
      • Study Guide
      • What you'll learn
      • Custom Actions Intro
      • Memory and Slots
      • Rasa Forms
      • Ask for Data
      • Data Validation
      • Example Form: Names
      • Slots
      • Custom Actions
      • Basic Forms
      • Custom Forms
  • Home
    • What you will learn
    • Introduction to CALM
    • Architecture of an Assistant built with CALM
    • Configuring the Working Environment
    • Difining Business Logic (Flows)
    • Dialogue Understanding
    • Custom Actions
    • Handling Deviations from the Happy Path
    • Enterprise Search
    • Running in Production
    • Summary
    • What you'll learn
    • Word Embeddings
    • Transformers
    • Tokenization
    • Bert
    • Shrinking Transformers
    • Model Confidence
    • Stopwords
    • Intents
    • End to End ML
    • GPT-3
    • Non-English Pipelines
    • Multi-Lingual Pipelines
    • Annotating Language Data
    • Featurizers
    • Regular Expressions
    • Transfer-Learning
    • What you'll learn
    • Measuring Bias
    • Mitigating Bias
    • Lipstick on a Pig
    • Projection Maths
    • Word Analogies
    • What you'll learn
    • Self Attention
    • Keys, Values & Queries
    • Multi Head Attention
    • Transformers
    • Overview
      • Introduction
      • Introduction to Rasa
      • Creating a New Assistant
      • The Domain File
      • Training Data & Rules
      • Entities
      • Slots
      • Responses
      • Pipeline
      • Custom Actions
      • Basic Forms
      • Custom Forms
      • Website Integration
      • Rasa is Always Improving
      • Windows
      • Mac
      • Ubuntu
      • What you'll learn
      • Custom Actions
      • Introduction to Slots
      • Constraining Slots
      • Rasa Forms
      • Text Slots
      • Data Validation
      • What you'll Learn
      • Introduction to Docker
      • Premade Rasa Containers
      • The Need for Kubernetes
      • How Kubernetes Works
      • Local Kubernetes Demo
      • Why Helm
      • Study Guide
      • What you'll learn
      • Custom Actions Intro
      • Memory and Slots
      • Rasa Forms
      • Ask for Data
      • Data Validation
      • Example Form: Names
      • Slots
      • Custom Actions
      • Basic Forms
      • Custom Forms

NLP for Developers: BERT

In this video, Rasa Senior Developer Advocate Rachael will talk about BERT, how it works and some of the benefits and drawbacks of the method.

Video#


Links#

  • Compressing BERT for faster prediction, Sam Sucik
  • Rasa LanguageModelFeaturizer Component
Prev
Next
2016-2024 © Rasa.