SlideShare uma empresa Scribd logo
1 de 16
Baixar para ler offline
Benchmarking approaches to
transfer learning in NLP
Yury Kashnitskiy, Gianluigi Bardelloni
Benchmarking approaches to
transfer learning in NLP
General problem:
1. Scarce well-labeled data in NLP tasks
2. Loads of raw texts available
3. How to utilize raw data to improve
performance in supervised NLP tasks?
Simpler: how to use all these bulks of unlabeled texts?
2
Business problem:
1. NLP tasks typically require a lot of labeled data
2. Labeling data is expensive
3. A disciplined approach to minimizing the needed training
size is highly desired
Benchmarking approaches to
transfer learning in NLP
3
Background: transfer learning in Computer Vision
Benchmarking approaches to
transfer learning in NLP
Pretraining
Fine-tuning
4
General idea:
1. A neural net is trained on raw
data to predict a word given its
context
2. Meanwhile it learns a vector
(“embedding”) for each word
3. Embeddings can be
transferred and used in
supervised learning tasks
(ex: part-of-speech-tagging)
This approach leads to SotA
results in many NLP tasks
Benchmarking approaches to
transfer learning in NLP
5
Benchmarking approaches to
transfer learning in NLP
ULMFiT:
1. Take a pretrained
language Model (with
ex. Wikipedia)
2. Fine-tune language
model on your domain
(ex. chats with
customers)
3. Fine-tune classifier
6
Benchmarking approaches to
transfer learning in NLP
Task 1. Amazon product reviews classification (English)
Validation accuracy:
Logistic Regression + Tf-Idf: 72.5%
ULMFiT: 79.5%
7
Benchmarking approaches to
transfer learning in NLP
Interpreting Logistic Regression with eli5
8
Examples of generated text (pure LSTM, PyTorch examples):
"This is a product that gets great secrets to the store's difficulty,
though it's compact and well made. Ordered one in the
mornings, for price. It's good though. Lasts 6 months, really need
to ask Amazon.com) as soon as you have the off!"
"I have been using Panasonic for a year and my mom was on
them to replace my Quadra Action and been pleased with the
gel.”
9
Benchmarking approaches to
transfer learning in NLP
Benchmarking approaches to
transfer learning in NLP
Task 2. Classifying chats with customers (Dutch)
Validation accuracy:
Logistic Regression + Tf-Idf: 73.5%
ULMFiT: 70.2%
Logit + ELMo: 66%
10
Benchmarking approaches to
transfer learning in NLP
Things to try:
1. Training ULMFiT
models with Dutch
texts
2. Fine-tuning BERT
classifier
3. Trying other
models: GPT-2,
OpenAI transformers
etc.
The goal is to develop best practices for
transfer learning in Dutch classification tasks
11
Benchmarking approaches to
transfer learning in NLP
What we expect from the collaboration:
1. Trying different transfer learning approaches
2. Both public and private data (English and Dutch)
3. Sharing code & ideas
In 3 months:
share preliminary results – code, models, guides
12
Proposed tasks
1. Benchmarking different approaches on several datasets to
see what works best. The main focus is on BERT and ULMFiT,
however, no limitations
Kaggle Dataset:
- Amazon healthcare reviews (English)
- Amazon pet products reviews (English), Kaggle comp.
- Clickbait news detection (English), Kaggle comp.
- Book reviews sentiment prediction (Dutch)
13
Benchmarking approaches to
transfer learning in NLP
2. Training own models for ULMFiT (Dutch)
3. Exploring Byte Pair Encoding as preprocessing for ULMFiT
4. Exploring preprocessing steps to improve BERT classifier
5. Dealing with typos and noise in text in case of BERT
6. Fine-tuning BERT Language models, exploring it's effect on
classification
14
Proposed tasks
Benchmarking approaches to
transfer learning in NLP
Trying other approaches:
1. Huggingface transfer learning tutorial + code
2. Fine-tuning classification head over LSTMs (pure Python)
3. GPT-2 transformers
4. Other OpenAI transformers
Other tasks:
1. Investigating text augmentations and their effect on
classification accuracy
2. Active learning in NLP
3. Hierarchical text classification
4. Few-shot learning (ex. Unsupervised Data Augmentation)
15
Benchmarking approaches to
transfer learning in NLP
Yury Kashnitskiy, Gianluigi Bardelloni
yury.kashnitskiy@kpn.com
gianluigi.bardelloni@kpn.com
16

Mais conteúdo relacionado

Mais procurados

Nlp presentation
Nlp presentationNlp presentation
Nlp presentationSurya Sg
 
Nautral Langauge Processing - Basics / Non Technical
Nautral Langauge Processing - Basics / Non Technical Nautral Langauge Processing - Basics / Non Technical
Nautral Langauge Processing - Basics / Non Technical Dhruv Gohil
 
GENETIC APPROACH FOR ARABIC PART OF SPEECH TAGGING
GENETIC APPROACH FOR ARABIC PART OF SPEECH TAGGINGGENETIC APPROACH FOR ARABIC PART OF SPEECH TAGGING
GENETIC APPROACH FOR ARABIC PART OF SPEECH TAGGINGijnlc
 
NLP and its application in Insurance -Short story presentation
NLP and its application in Insurance -Short story presentationNLP and its application in Insurance -Short story presentation
NLP and its application in Insurance -Short story presentationstuti_agarwal
 
Bt0081, software engineering
Bt0081, software engineeringBt0081, software engineering
Bt0081, software engineeringsmumbahelp
 
Bt0081, software engineering
Bt0081, software engineeringBt0081, software engineering
Bt0081, software engineeringsmumbahelp
 
A neural probabilistic language model
A neural probabilistic language modelA neural probabilistic language model
A neural probabilistic language modelc sharada
 
Monte Carlo Modelling of Confidence Intervals in Translation Quality Evaluati...
Monte Carlo Modelling of Confidence Intervals in Translation Quality Evaluati...Monte Carlo Modelling of Confidence Intervals in Translation Quality Evaluati...
Monte Carlo Modelling of Confidence Intervals in Translation Quality Evaluati...Lifeng (Aaron) Han
 
Next word predication using markov models
Next word predication using markov modelsNext word predication using markov models
Next word predication using markov modelsREZAUL KARIM REFATH
 
LEPOR: an augmented machine translation evaluation metric - Thesis PPT
LEPOR: an augmented machine translation evaluation metric - Thesis PPT LEPOR: an augmented machine translation evaluation metric - Thesis PPT
LEPOR: an augmented machine translation evaluation metric - Thesis PPT Lifeng (Aaron) Han
 
MT and Post Editing in master's level translation education
MT and Post Editing in master's level translation education MT and Post Editing in master's level translation education
MT and Post Editing in master's level translation education Jakub Absolon
 
ESR8 Liangyou Li - EXPERT Summer School - Malaga 2015
ESR8 Liangyou Li - EXPERT Summer School - Malaga 2015ESR8 Liangyou Li - EXPERT Summer School - Malaga 2015
ESR8 Liangyou Li - EXPERT Summer School - Malaga 2015RIILP
 
Seq2seq Model to Tokenize the Chinese Language
Seq2seq Model to Tokenize the Chinese LanguageSeq2seq Model to Tokenize the Chinese Language
Seq2seq Model to Tokenize the Chinese LanguageJinho Choi
 
NLP using transformers
NLP using transformers NLP using transformers
NLP using transformers Arvind Devaraj
 
GENERATING SUMMARIES USING SENTENCE COMPRESSION AND STATISTICAL MEASURES
GENERATING SUMMARIES USING SENTENCE COMPRESSION AND STATISTICAL MEASURESGENERATING SUMMARIES USING SENTENCE COMPRESSION AND STATISTICAL MEASURES
GENERATING SUMMARIES USING SENTENCE COMPRESSION AND STATISTICAL MEASURESijnlc
 
COLING 2014: Joint Opinion Relation Detection Using One-Class Deep Neural Net...
COLING 2014: Joint Opinion Relation Detection Using One-Class Deep Neural Net...COLING 2014: Joint Opinion Relation Detection Using One-Class Deep Neural Net...
COLING 2014: Joint Opinion Relation Detection Using One-Class Deep Neural Net...Peinan ZHANG
 
SEMI-AUTOMATIC SIMULTANEOUS INTERPRETING QUALITY EVALUATION
SEMI-AUTOMATIC SIMULTANEOUS INTERPRETING QUALITY EVALUATIONSEMI-AUTOMATIC SIMULTANEOUS INTERPRETING QUALITY EVALUATION
SEMI-AUTOMATIC SIMULTANEOUS INTERPRETING QUALITY EVALUATIONijnlc
 

Mais procurados (20)

Nlp presentation
Nlp presentationNlp presentation
Nlp presentation
 
Nautral Langauge Processing - Basics / Non Technical
Nautral Langauge Processing - Basics / Non Technical Nautral Langauge Processing - Basics / Non Technical
Nautral Langauge Processing - Basics / Non Technical
 
Plug play language_models
Plug play language_modelsPlug play language_models
Plug play language_models
 
GENETIC APPROACH FOR ARABIC PART OF SPEECH TAGGING
GENETIC APPROACH FOR ARABIC PART OF SPEECH TAGGINGGENETIC APPROACH FOR ARABIC PART OF SPEECH TAGGING
GENETIC APPROACH FOR ARABIC PART OF SPEECH TAGGING
 
NLP and its application in Insurance -Short story presentation
NLP and its application in Insurance -Short story presentationNLP and its application in Insurance -Short story presentation
NLP and its application in Insurance -Short story presentation
 
Bt0081, software engineering
Bt0081, software engineeringBt0081, software engineering
Bt0081, software engineering
 
Bt0081, software engineering
Bt0081, software engineeringBt0081, software engineering
Bt0081, software engineering
 
A neural probabilistic language model
A neural probabilistic language modelA neural probabilistic language model
A neural probabilistic language model
 
Monte Carlo Modelling of Confidence Intervals in Translation Quality Evaluati...
Monte Carlo Modelling of Confidence Intervals in Translation Quality Evaluati...Monte Carlo Modelling of Confidence Intervals in Translation Quality Evaluati...
Monte Carlo Modelling of Confidence Intervals in Translation Quality Evaluati...
 
Next word predication using markov models
Next word predication using markov modelsNext word predication using markov models
Next word predication using markov models
 
LEPOR: an augmented machine translation evaluation metric - Thesis PPT
LEPOR: an augmented machine translation evaluation metric - Thesis PPT LEPOR: an augmented machine translation evaluation metric - Thesis PPT
LEPOR: an augmented machine translation evaluation metric - Thesis PPT
 
MT and Post Editing in master's level translation education
MT and Post Editing in master's level translation education MT and Post Editing in master's level translation education
MT and Post Editing in master's level translation education
 
Blenderbot
BlenderbotBlenderbot
Blenderbot
 
ESR8 Liangyou Li - EXPERT Summer School - Malaga 2015
ESR8 Liangyou Li - EXPERT Summer School - Malaga 2015ESR8 Liangyou Li - EXPERT Summer School - Malaga 2015
ESR8 Liangyou Li - EXPERT Summer School - Malaga 2015
 
Seq2seq Model to Tokenize the Chinese Language
Seq2seq Model to Tokenize the Chinese LanguageSeq2seq Model to Tokenize the Chinese Language
Seq2seq Model to Tokenize the Chinese Language
 
NLP using transformers
NLP using transformers NLP using transformers
NLP using transformers
 
GENERATING SUMMARIES USING SENTENCE COMPRESSION AND STATISTICAL MEASURES
GENERATING SUMMARIES USING SENTENCE COMPRESSION AND STATISTICAL MEASURESGENERATING SUMMARIES USING SENTENCE COMPRESSION AND STATISTICAL MEASURES
GENERATING SUMMARIES USING SENTENCE COMPRESSION AND STATISTICAL MEASURES
 
Tensorflow
TensorflowTensorflow
Tensorflow
 
COLING 2014: Joint Opinion Relation Detection Using One-Class Deep Neural Net...
COLING 2014: Joint Opinion Relation Detection Using One-Class Deep Neural Net...COLING 2014: Joint Opinion Relation Detection Using One-Class Deep Neural Net...
COLING 2014: Joint Opinion Relation Detection Using One-Class Deep Neural Net...
 
SEMI-AUTOMATIC SIMULTANEOUS INTERPRETING QUALITY EVALUATION
SEMI-AUTOMATIC SIMULTANEOUS INTERPRETING QUALITY EVALUATIONSEMI-AUTOMATIC SIMULTANEOUS INTERPRETING QUALITY EVALUATION
SEMI-AUTOMATIC SIMULTANEOUS INTERPRETING QUALITY EVALUATION
 

Semelhante a Benchmarking transfer learning approaches for NLP

Thomas Wolf "An Introduction to Transfer Learning and Hugging Face"
Thomas Wolf "An Introduction to Transfer Learning and Hugging Face"Thomas Wolf "An Introduction to Transfer Learning and Hugging Face"
Thomas Wolf "An Introduction to Transfer Learning and Hugging Face"Fwdays
 
Thomas Wolf "Transfer learning in NLP"
Thomas Wolf "Transfer learning in NLP"Thomas Wolf "Transfer learning in NLP"
Thomas Wolf "Transfer learning in NLP"Fwdays
 
Training language models to follow instructions with human feedback (Instruct...
Training language models to follow instructions with human feedback (Instruct...Training language models to follow instructions with human feedback (Instruct...
Training language models to follow instructions with human feedback (Instruct...Rama Irsheidat
 
A Comparative Study of Text Comprehension in IELTS Reading Exam using GPT-3
A Comparative Study of Text Comprehension in IELTS Reading Exam using GPT-3A Comparative Study of Text Comprehension in IELTS Reading Exam using GPT-3
A Comparative Study of Text Comprehension in IELTS Reading Exam using GPT-3AIRCC Publishing Corporation
 
Deep Neural Networks in Text Classification using Active Learning
Deep Neural Networks in Text Classification using Active LearningDeep Neural Networks in Text Classification using Active Learning
Deep Neural Networks in Text Classification using Active LearningMirsaeid Abolghasemi
 
Promise 2011: "An Iterative Semi-supervised Approach to Software Fault Predic...
Promise 2011: "An Iterative Semi-supervised Approach to Software Fault Predic...Promise 2011: "An Iterative Semi-supervised Approach to Software Fault Predic...
Promise 2011: "An Iterative Semi-supervised Approach to Software Fault Predic...CS, NcState
 
NLP Meetup 2023
NLP Meetup 2023NLP Meetup 2023
NLP Meetup 2023GabiMaeztu
 
Natural Language Generation / Stanford cs224n 2019w lecture 15 Review
Natural Language Generation / Stanford cs224n 2019w lecture 15 ReviewNatural Language Generation / Stanford cs224n 2019w lecture 15 Review
Natural Language Generation / Stanford cs224n 2019w lecture 15 Reviewchangedaeoh
 
Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot T...
Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot T...Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot T...
Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot T...eraser Juan José Calderón
 
Natural Language Processing: From Human-Robot Interaction to Alzheimer’s Dete...
Natural Language Processing: From Human-Robot Interaction to Alzheimer’s Dete...Natural Language Processing: From Human-Robot Interaction to Alzheimer’s Dete...
Natural Language Processing: From Human-Robot Interaction to Alzheimer’s Dete...Jekaterina Novikova, PhD
 
Deep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
Deep Learning for NLP (without Magic) - Richard Socher and Christopher ManningDeep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
Deep Learning for NLP (without Magic) - Richard Socher and Christopher ManningBigDataCloud
 
IRJET- Survey on Deep Learning Approaches for Phrase Structure Identification...
IRJET- Survey on Deep Learning Approaches for Phrase Structure Identification...IRJET- Survey on Deep Learning Approaches for Phrase Structure Identification...
IRJET- Survey on Deep Learning Approaches for Phrase Structure Identification...IRJET Journal
 
How to fine-tune and develop your own large language model.pptx
How to fine-tune and develop your own large language model.pptxHow to fine-tune and develop your own large language model.pptx
How to fine-tune and develop your own large language model.pptxKnoldus Inc.
 
Implications of GPT-3
Implications of GPT-3Implications of GPT-3
Implications of GPT-3Raven Jiang
 
Scaling Instruction-Finetuned Language Models
Scaling Instruction-Finetuned Language ModelsScaling Instruction-Finetuned Language Models
Scaling Instruction-Finetuned Language Modelstaeseon ryu
 
Transfer_Learning_for_Natural_Language_P_v3_MEAP.pdf
Transfer_Learning_for_Natural_Language_P_v3_MEAP.pdfTransfer_Learning_for_Natural_Language_P_v3_MEAP.pdf
Transfer_Learning_for_Natural_Language_P_v3_MEAP.pdforanisalcani
 
Natural language processing for requirements engineering: ICSE 2021 Technical...
Natural language processing for requirements engineering: ICSE 2021 Technical...Natural language processing for requirements engineering: ICSE 2021 Technical...
Natural language processing for requirements engineering: ICSE 2021 Technical...alessio_ferrari
 

Semelhante a Benchmarking transfer learning approaches for NLP (20)

Thomas Wolf "An Introduction to Transfer Learning and Hugging Face"
Thomas Wolf "An Introduction to Transfer Learning and Hugging Face"Thomas Wolf "An Introduction to Transfer Learning and Hugging Face"
Thomas Wolf "An Introduction to Transfer Learning and Hugging Face"
 
Thomas Wolf "Transfer learning in NLP"
Thomas Wolf "Transfer learning in NLP"Thomas Wolf "Transfer learning in NLP"
Thomas Wolf "Transfer learning in NLP"
 
Training language models to follow instructions with human feedback (Instruct...
Training language models to follow instructions with human feedback (Instruct...Training language models to follow instructions with human feedback (Instruct...
Training language models to follow instructions with human feedback (Instruct...
 
A Comparative Study of Text Comprehension in IELTS Reading Exam using GPT-3
A Comparative Study of Text Comprehension in IELTS Reading Exam using GPT-3A Comparative Study of Text Comprehension in IELTS Reading Exam using GPT-3
A Comparative Study of Text Comprehension in IELTS Reading Exam using GPT-3
 
Deep Neural Networks in Text Classification using Active Learning
Deep Neural Networks in Text Classification using Active LearningDeep Neural Networks in Text Classification using Active Learning
Deep Neural Networks in Text Classification using Active Learning
 
Promise 2011: "An Iterative Semi-supervised Approach to Software Fault Predic...
Promise 2011: "An Iterative Semi-supervised Approach to Software Fault Predic...Promise 2011: "An Iterative Semi-supervised Approach to Software Fault Predic...
Promise 2011: "An Iterative Semi-supervised Approach to Software Fault Predic...
 
srinu.pptx
srinu.pptxsrinu.pptx
srinu.pptx
 
NLP Meetup 2023
NLP Meetup 2023NLP Meetup 2023
NLP Meetup 2023
 
2201.00598.pdf
2201.00598.pdf2201.00598.pdf
2201.00598.pdf
 
Natural Language Generation / Stanford cs224n 2019w lecture 15 Review
Natural Language Generation / Stanford cs224n 2019w lecture 15 ReviewNatural Language Generation / Stanford cs224n 2019w lecture 15 Review
Natural Language Generation / Stanford cs224n 2019w lecture 15 Review
 
Cd project
Cd projectCd project
Cd project
 
Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot T...
Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot T...Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot T...
Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot T...
 
Natural Language Processing: From Human-Robot Interaction to Alzheimer’s Dete...
Natural Language Processing: From Human-Robot Interaction to Alzheimer’s Dete...Natural Language Processing: From Human-Robot Interaction to Alzheimer’s Dete...
Natural Language Processing: From Human-Robot Interaction to Alzheimer’s Dete...
 
Deep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
Deep Learning for NLP (without Magic) - Richard Socher and Christopher ManningDeep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
Deep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
 
IRJET- Survey on Deep Learning Approaches for Phrase Structure Identification...
IRJET- Survey on Deep Learning Approaches for Phrase Structure Identification...IRJET- Survey on Deep Learning Approaches for Phrase Structure Identification...
IRJET- Survey on Deep Learning Approaches for Phrase Structure Identification...
 
How to fine-tune and develop your own large language model.pptx
How to fine-tune and develop your own large language model.pptxHow to fine-tune and develop your own large language model.pptx
How to fine-tune and develop your own large language model.pptx
 
Implications of GPT-3
Implications of GPT-3Implications of GPT-3
Implications of GPT-3
 
Scaling Instruction-Finetuned Language Models
Scaling Instruction-Finetuned Language ModelsScaling Instruction-Finetuned Language Models
Scaling Instruction-Finetuned Language Models
 
Transfer_Learning_for_Natural_Language_P_v3_MEAP.pdf
Transfer_Learning_for_Natural_Language_P_v3_MEAP.pdfTransfer_Learning_for_Natural_Language_P_v3_MEAP.pdf
Transfer_Learning_for_Natural_Language_P_v3_MEAP.pdf
 
Natural language processing for requirements engineering: ICSE 2021 Technical...
Natural language processing for requirements engineering: ICSE 2021 Technical...Natural language processing for requirements engineering: ICSE 2021 Technical...
Natural language processing for requirements engineering: ICSE 2021 Technical...
 

Mais de Yury Kashnitsky

How to jump into Data Science
How to jump into Data ScienceHow to jump into Data Science
How to jump into Data ScienceYury Kashnitsky
 
mlcourse.ai fall2019 Live Session 0
mlcourse.ai fall2019 Live Session 0mlcourse.ai fall2019 Live Session 0
mlcourse.ai fall2019 Live Session 0Yury Kashnitsky
 
Gender-unbiased BERT-based Pronoun Resolution
Gender-unbiased BERT-based  Pronoun ResolutionGender-unbiased BERT-based  Pronoun Resolution
Gender-unbiased BERT-based Pronoun ResolutionYury Kashnitsky
 
Time series forecasting with ARIMA
Time series forecasting with ARIMATime series forecasting with ARIMA
Time series forecasting with ARIMAYury Kashnitsky
 
mlcourse.ai, introduction, course overview
mlcourse.ai, introduction, course overviewmlcourse.ai, introduction, course overview
mlcourse.ai, introduction, course overviewYury Kashnitsky
 
Необычные модели Playboy, или про поиск аномалий в данных
Необычные модели Playboy, или про поиск аномалий в данныхНеобычные модели Playboy, или про поиск аномалий в данных
Необычные модели Playboy, или про поиск аномалий в данныхYury Kashnitsky
 

Mais de Yury Kashnitsky (8)

How to jump into Data Science
How to jump into Data ScienceHow to jump into Data Science
How to jump into Data Science
 
mlcourse.ai fall2019 Live Session 0
mlcourse.ai fall2019 Live Session 0mlcourse.ai fall2019 Live Session 0
mlcourse.ai fall2019 Live Session 0
 
Gender-unbiased BERT-based Pronoun Resolution
Gender-unbiased BERT-based  Pronoun ResolutionGender-unbiased BERT-based  Pronoun Resolution
Gender-unbiased BERT-based Pronoun Resolution
 
mlcourse.ai. Outro
mlcourse.ai. Outromlcourse.ai. Outro
mlcourse.ai. Outro
 
Time series forecasting with ARIMA
Time series forecasting with ARIMATime series forecasting with ARIMA
Time series forecasting with ARIMA
 
mlcourse.ai. Clustering
mlcourse.ai. Clusteringmlcourse.ai. Clustering
mlcourse.ai. Clustering
 
mlcourse.ai, introduction, course overview
mlcourse.ai, introduction, course overviewmlcourse.ai, introduction, course overview
mlcourse.ai, introduction, course overview
 
Необычные модели Playboy, или про поиск аномалий в данных
Необычные модели Playboy, или про поиск аномалий в данныхНеобычные модели Playboy, или про поиск аномалий в данных
Необычные модели Playboy, или про поиск аномалий в данных
 

Último

Semantic Shed - Squashing and Squeezing.pptx
Semantic Shed - Squashing and Squeezing.pptxSemantic Shed - Squashing and Squeezing.pptx
Semantic Shed - Squashing and Squeezing.pptxMike Bennett
 
Top 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In QueensTop 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In Queensdataanalyticsqueen03
 
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一F sss
 
Real-Time AI Streaming - AI Max Princeton
Real-Time AI  Streaming - AI Max PrincetonReal-Time AI  Streaming - AI Max Princeton
Real-Time AI Streaming - AI Max PrincetonTimothy Spann
 
Conf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
Conf42-LLM_Adding Generative AI to Real-Time Streaming PipelinesConf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
Conf42-LLM_Adding Generative AI to Real-Time Streaming PipelinesTimothy Spann
 
RABBIT: A CLI tool for identifying bots based on their GitHub events.
RABBIT: A CLI tool for identifying bots based on their GitHub events.RABBIT: A CLI tool for identifying bots based on their GitHub events.
RABBIT: A CLI tool for identifying bots based on their GitHub events.natarajan8993
 
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...Boston Institute of Analytics
 
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)jennyeacort
 
Identifying Appropriate Test Statistics Involving Population Mean
Identifying Appropriate Test Statistics Involving Population MeanIdentifying Appropriate Test Statistics Involving Population Mean
Identifying Appropriate Test Statistics Involving Population MeanMYRABACSAFRA2
 
Biometric Authentication: The Evolution, Applications, Benefits and Challenge...
Biometric Authentication: The Evolution, Applications, Benefits and Challenge...Biometric Authentication: The Evolution, Applications, Benefits and Challenge...
Biometric Authentication: The Evolution, Applications, Benefits and Challenge...GQ Research
 
LLMs, LMMs, their Improvement Suggestions and the Path towards AGI
LLMs, LMMs, their Improvement Suggestions and the Path towards AGILLMs, LMMs, their Improvement Suggestions and the Path towards AGI
LLMs, LMMs, their Improvement Suggestions and the Path towards AGIThomas Poetter
 
How we prevented account sharing with MFA
How we prevented account sharing with MFAHow we prevented account sharing with MFA
How we prevented account sharing with MFAAndrei Kaleshka
 
Minimizing AI Hallucinations/Confabulations and the Path towards AGI with Exa...
Minimizing AI Hallucinations/Confabulations and the Path towards AGI with Exa...Minimizing AI Hallucinations/Confabulations and the Path towards AGI with Exa...
Minimizing AI Hallucinations/Confabulations and the Path towards AGI with Exa...Thomas Poetter
 
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort servicejennyeacort
 
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...Boston Institute of Analytics
 
Predicting Salary Using Data Science: A Comprehensive Analysis.pdf
Predicting Salary Using Data Science: A Comprehensive Analysis.pdfPredicting Salary Using Data Science: A Comprehensive Analysis.pdf
Predicting Salary Using Data Science: A Comprehensive Analysis.pdfBoston Institute of Analytics
 
办美国阿肯色大学小石城分校毕业证成绩单pdf电子版制作修改#真实留信入库#永久存档#真实可查#diploma#degree
办美国阿肯色大学小石城分校毕业证成绩单pdf电子版制作修改#真实留信入库#永久存档#真实可查#diploma#degree办美国阿肯色大学小石城分校毕业证成绩单pdf电子版制作修改#真实留信入库#永久存档#真实可查#diploma#degree
办美国阿肯色大学小石城分校毕业证成绩单pdf电子版制作修改#真实留信入库#永久存档#真实可查#diploma#degreeyuu sss
 
NO1 Certified Black Magic Specialist Expert Amil baba in Lahore Islamabad Raw...
NO1 Certified Black Magic Specialist Expert Amil baba in Lahore Islamabad Raw...NO1 Certified Black Magic Specialist Expert Amil baba in Lahore Islamabad Raw...
NO1 Certified Black Magic Specialist Expert Amil baba in Lahore Islamabad Raw...Amil Baba Dawood bangali
 
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档208367051
 

Último (20)

Semantic Shed - Squashing and Squeezing.pptx
Semantic Shed - Squashing and Squeezing.pptxSemantic Shed - Squashing and Squeezing.pptx
Semantic Shed - Squashing and Squeezing.pptx
 
Top 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In QueensTop 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In Queens
 
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
 
Real-Time AI Streaming - AI Max Princeton
Real-Time AI  Streaming - AI Max PrincetonReal-Time AI  Streaming - AI Max Princeton
Real-Time AI Streaming - AI Max Princeton
 
Deep Generative Learning for All - The Gen AI Hype (Spring 2024)
Deep Generative Learning for All - The Gen AI Hype (Spring 2024)Deep Generative Learning for All - The Gen AI Hype (Spring 2024)
Deep Generative Learning for All - The Gen AI Hype (Spring 2024)
 
Conf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
Conf42-LLM_Adding Generative AI to Real-Time Streaming PipelinesConf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
Conf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
 
RABBIT: A CLI tool for identifying bots based on their GitHub events.
RABBIT: A CLI tool for identifying bots based on their GitHub events.RABBIT: A CLI tool for identifying bots based on their GitHub events.
RABBIT: A CLI tool for identifying bots based on their GitHub events.
 
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
 
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
 
Identifying Appropriate Test Statistics Involving Population Mean
Identifying Appropriate Test Statistics Involving Population MeanIdentifying Appropriate Test Statistics Involving Population Mean
Identifying Appropriate Test Statistics Involving Population Mean
 
Biometric Authentication: The Evolution, Applications, Benefits and Challenge...
Biometric Authentication: The Evolution, Applications, Benefits and Challenge...Biometric Authentication: The Evolution, Applications, Benefits and Challenge...
Biometric Authentication: The Evolution, Applications, Benefits and Challenge...
 
LLMs, LMMs, their Improvement Suggestions and the Path towards AGI
LLMs, LMMs, their Improvement Suggestions and the Path towards AGILLMs, LMMs, their Improvement Suggestions and the Path towards AGI
LLMs, LMMs, their Improvement Suggestions and the Path towards AGI
 
How we prevented account sharing with MFA
How we prevented account sharing with MFAHow we prevented account sharing with MFA
How we prevented account sharing with MFA
 
Minimizing AI Hallucinations/Confabulations and the Path towards AGI with Exa...
Minimizing AI Hallucinations/Confabulations and the Path towards AGI with Exa...Minimizing AI Hallucinations/Confabulations and the Path towards AGI with Exa...
Minimizing AI Hallucinations/Confabulations and the Path towards AGI with Exa...
 
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
 
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
 
Predicting Salary Using Data Science: A Comprehensive Analysis.pdf
Predicting Salary Using Data Science: A Comprehensive Analysis.pdfPredicting Salary Using Data Science: A Comprehensive Analysis.pdf
Predicting Salary Using Data Science: A Comprehensive Analysis.pdf
 
办美国阿肯色大学小石城分校毕业证成绩单pdf电子版制作修改#真实留信入库#永久存档#真实可查#diploma#degree
办美国阿肯色大学小石城分校毕业证成绩单pdf电子版制作修改#真实留信入库#永久存档#真实可查#diploma#degree办美国阿肯色大学小石城分校毕业证成绩单pdf电子版制作修改#真实留信入库#永久存档#真实可查#diploma#degree
办美国阿肯色大学小石城分校毕业证成绩单pdf电子版制作修改#真实留信入库#永久存档#真实可查#diploma#degree
 
NO1 Certified Black Magic Specialist Expert Amil baba in Lahore Islamabad Raw...
NO1 Certified Black Magic Specialist Expert Amil baba in Lahore Islamabad Raw...NO1 Certified Black Magic Specialist Expert Amil baba in Lahore Islamabad Raw...
NO1 Certified Black Magic Specialist Expert Amil baba in Lahore Islamabad Raw...
 
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
 

Benchmarking transfer learning approaches for NLP

  • 1. Benchmarking approaches to transfer learning in NLP Yury Kashnitskiy, Gianluigi Bardelloni
  • 2. Benchmarking approaches to transfer learning in NLP General problem: 1. Scarce well-labeled data in NLP tasks 2. Loads of raw texts available 3. How to utilize raw data to improve performance in supervised NLP tasks? Simpler: how to use all these bulks of unlabeled texts? 2
  • 3. Business problem: 1. NLP tasks typically require a lot of labeled data 2. Labeling data is expensive 3. A disciplined approach to minimizing the needed training size is highly desired Benchmarking approaches to transfer learning in NLP 3
  • 4. Background: transfer learning in Computer Vision Benchmarking approaches to transfer learning in NLP Pretraining Fine-tuning 4
  • 5. General idea: 1. A neural net is trained on raw data to predict a word given its context 2. Meanwhile it learns a vector (“embedding”) for each word 3. Embeddings can be transferred and used in supervised learning tasks (ex: part-of-speech-tagging) This approach leads to SotA results in many NLP tasks Benchmarking approaches to transfer learning in NLP 5
  • 6. Benchmarking approaches to transfer learning in NLP ULMFiT: 1. Take a pretrained language Model (with ex. Wikipedia) 2. Fine-tune language model on your domain (ex. chats with customers) 3. Fine-tune classifier 6
  • 7. Benchmarking approaches to transfer learning in NLP Task 1. Amazon product reviews classification (English) Validation accuracy: Logistic Regression + Tf-Idf: 72.5% ULMFiT: 79.5% 7
  • 8. Benchmarking approaches to transfer learning in NLP Interpreting Logistic Regression with eli5 8
  • 9. Examples of generated text (pure LSTM, PyTorch examples): "This is a product that gets great secrets to the store's difficulty, though it's compact and well made. Ordered one in the mornings, for price. It's good though. Lasts 6 months, really need to ask Amazon.com) as soon as you have the off!" "I have been using Panasonic for a year and my mom was on them to replace my Quadra Action and been pleased with the gel.” 9 Benchmarking approaches to transfer learning in NLP
  • 10. Benchmarking approaches to transfer learning in NLP Task 2. Classifying chats with customers (Dutch) Validation accuracy: Logistic Regression + Tf-Idf: 73.5% ULMFiT: 70.2% Logit + ELMo: 66% 10
  • 11. Benchmarking approaches to transfer learning in NLP Things to try: 1. Training ULMFiT models with Dutch texts 2. Fine-tuning BERT classifier 3. Trying other models: GPT-2, OpenAI transformers etc. The goal is to develop best practices for transfer learning in Dutch classification tasks 11
  • 12. Benchmarking approaches to transfer learning in NLP What we expect from the collaboration: 1. Trying different transfer learning approaches 2. Both public and private data (English and Dutch) 3. Sharing code & ideas In 3 months: share preliminary results – code, models, guides 12
  • 13. Proposed tasks 1. Benchmarking different approaches on several datasets to see what works best. The main focus is on BERT and ULMFiT, however, no limitations Kaggle Dataset: - Amazon healthcare reviews (English) - Amazon pet products reviews (English), Kaggle comp. - Clickbait news detection (English), Kaggle comp. - Book reviews sentiment prediction (Dutch) 13 Benchmarking approaches to transfer learning in NLP
  • 14. 2. Training own models for ULMFiT (Dutch) 3. Exploring Byte Pair Encoding as preprocessing for ULMFiT 4. Exploring preprocessing steps to improve BERT classifier 5. Dealing with typos and noise in text in case of BERT 6. Fine-tuning BERT Language models, exploring it's effect on classification 14 Proposed tasks Benchmarking approaches to transfer learning in NLP
  • 15. Trying other approaches: 1. Huggingface transfer learning tutorial + code 2. Fine-tuning classification head over LSTMs (pure Python) 3. GPT-2 transformers 4. Other OpenAI transformers Other tasks: 1. Investigating text augmentations and their effect on classification accuracy 2. Active learning in NLP 3. Hierarchical text classification 4. Few-shot learning (ex. Unsupervised Data Augmentation) 15 Benchmarking approaches to transfer learning in NLP
  • 16. Yury Kashnitskiy, Gianluigi Bardelloni yury.kashnitskiy@kpn.com gianluigi.bardelloni@kpn.com 16