DEV Community

Natural Language Processing

NLP is Natural Language Processing the technology behind Home assistants and search engines.

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
Day 36: Fine-Tuning BERT for Downstream Tasks

Day 36: Fine-Tuning BERT for Downstream Tasks

Comments
5 min read
I Thought Extracting Dates Was Easy Until I Built an On-Device iOS App

I Thought Extracting Dates Was Easy Until I Built an On-Device iOS App

Comments
3 min read
What Changes When AI Text Is Humanized: 138 Pairs, Measured

What Changes When AI Text Is Humanized: 138 Pairs, Measured

Comments
11 min read
Was My 2,103-Token Tokenizer a Mistake?

Was My 2,103-Token Tokenizer a Mistake?

Comments
3 min read
ASR-induced failure rate: did the speech path break it?

ASR-induced failure rate: did the speech path break it?

Comments
5 min read
Day 35: BERT: Architecture and Masked Language Modeling

Day 35: BERT: Architecture and Masked Language Modeling

Comments
5 min read
How to Identify Technologies Mentioned in a YouTube Video

How to Identify Technologies Mentioned in a YouTube Video

Comments
2 min read
Evaluate AI Dialogue Rewrites by Tactic, Not Just Fluency

Evaluate AI Dialogue Rewrites by Tactic, Not Just Fluency

Comments
3 min read
Day 34: What Is Pretraining? Transfer Learning for NLP

Day 34: What Is Pretraining? Transfer Learning for NLP

Comments
4 min read
I Distilled a 568M Multilingual Model Into a 37M Japanese-English Encoder — Here's What Survived

I Distilled a 568M Multilingual Model Into a 37M Japanese-English Encoder — Here's What Survived

1
Comments
4 min read
Day 33: Building a Mini Transformer From Scratch (Code Walkthrough)

Day 33: Building a Mini Transformer From Scratch (Code Walkthrough)

Comments
5 min read
I trained a language model that writes by filling in blanks

I trained a language model that writes by filling in blanks

Comments
3 min read
Tokens Are a Crutch: Why Byte Models Win the Long Game

Tokens Are a Crutch: Why Byte Models Win the Long Game

Comments
5 min read
From Attention to a Working Language Model

From Attention to a Working Language Model

Comments 1
13 min read
Not every AI call needs to generate text: ej, an 11 MB model for typed decisions

Not every AI call needs to generate text: ej, an 11 MB model for typed decisions

Comments
6 min read
👋 Sign in for the ability to sort posts by relevant, latest, or top.