Description
Data Science: Transformers for Natural Language Processing, the name of the Data Science: Transformers for Natural Language Processing course published by the Udemy website. Welcome to the Data Science: Transformers for Natural Language Processing training. Deep learning hasn’t been the same since Transformers hit the scene. Machine learning is capable of producing text that is essentially indistinguishable from human-generated text. We have achieved advanced performance in many NLP tasks, such as machine translation, question answering, entailment, named entity recognition, and more. Am. We have created multi-faceted (text and image) templates that can create amazing works of art using just a text message. We have solved an old molecular biology problem called “protein structure prediction.” In this course you will learn very practical skills on using transformers and, if you wish, detailed theory on how transformers work and care. This differs from many other sources which only cover the former. This course is divided into 3 main parts:
- Use transformers
- Fine-tuning transformers
- Transformers in depth
Section 1: Using Transformers In this section you will learn how to use the transformers that you have been taught. It costs millions of dollars, so it’s not something you want to try for yourself! We will see how these pre-built models can be used for a wide range of tasks, including: text classification (e.g. spam detection, sentiment analysis, document classification) named entity recognition summarization translated text automatically, questions and answers, generation of (credible) text. ). Masked Language Modeling (Paper Rotation) Zero Shot Classification This is currently very practical. If you need to perform sentiment analysis, document categorization, entity detection, translation, summarization, etc. on documents at work or for your clients – now you have the most powerful advanced models with very little. Lines of Code is one of the most amazing “zero-shot classification” programs where you will see that a pre-trained model can classify your documents even without any training.
Part 2: Refine Transformers In this part, you will learn how to improve the performance of transformers in your custom dataset. With Transfer Learning, you can leverage the millions of dollars of training already spent keeping transformers running smoothly. You’ll find that you can fine-tune a transformer with relatively little work (and low cost). We’ll look at how to fine-tune transformers for the most practical real-world tasks, such as text classification (sentiment analysis, spam detection), entity detection, and machine translation.
Part 3: Transformers in Depth In this part, you will learn how transformers work. The first parts are nice, but a little too nice. Libraries are good for people who just want to work, but they won’t work if you want to do something new or interesting. Let’s be clear: it’s very practical. How practical is that, you might ask? Well, that’s where the big money is. Those who have a deep understanding of these models and can do things that no one has done before are able to achieve higher salaries and prestigious titles. Machine learning is a competitive field, and a deep understanding of how things work can be the edge you need to excel. We’ll also look at how to implement transformers from scratch. As the great Richard Feynman once said, “I don’t understand what I can’t create.” Suggested prerequisites: Adequate Python coding skills. Deep learning with CNN and RNN is useful but not required. Deep learning with Seq2Seq models is useful but not necessary for the in-depth section: Understanding the theory behind CNN, RNN and seq2seq is very useful. Updates to expect: More detailed tuning apps. More in-depth conceptual lectures. Transformers running from scratch.
Who should attend ?
- Anyone wishing to master natural language processing (NLP)
- Anyone who loves deep learning and wants to learn more about the most powerful neural networks (transformers)
- Anyone looking to go beyond the typical beginner-only courses on Udemy
What will you learn in Data Science: Transformative Courses for Natural Language Processing ,
- Apply transformers to real-world tasks with just a few lines of code
- Fine-tune transformers on your own datasets with transfer learning
- Sentiment analysis, spam detection, text classification
- NER (named entity recognition), part-of-speech tagging
- Create your own article spinner for SEO
- Generate credible, human-like text
- Neural machine translation and text summarization
- Answering questions (e.g. SQuAD)
- Zero shot rating
- Understanding Personal Attention and In-Depth Theory Behind Transformers
- Implement Transformers from Scratch
- Use Transformers with Tensorflow and PyTorch
- Understanding BERT, GPT, GPT-2, and GPT-3, and where to apply them
- Understanding encoder, decoder, and seq2seq architectures
- Master the Hugging Face Python library
Course details:
- Editor: Udemy
- Instructor: Lazy programmer team,Lazy Programmer Inc
- French language
- Training level: introduction to advanced
- Number of courses: 132
- Training duration: 18 hours 41 minutes
Course content 2023/10
Data Science: Transformers for Natural Language Processing Requirements ,
- Install Python, it’s free!
- Beginner and Intermediate Level Content: Decent Python Programming Skills
- Expert level content: Good understanding of CNNs and RNNs and ability to code in PyTorch or Tensorflow
Photo of the course:
Simple video:
installation guide ,
After ripping, view with your favorite player.
Subtitle: English
Quality: 720p
Changes:
Version 2023/3 compared to 2022/5 increased the number of 42 lessons and the duration by 5 hours and 40 minutes.
Version 2023/5 compared to 2023/3 increased the number by 1 lesson and the duration by 8 minutes.
Version 2023/8 compared to 2023/5 increased the number of 4 lessons and the duration by 40 minutes.
Download links:
File password(s): free download software
File size:
6.04 GB