Transfer Learning for Text Classification Using Hugging Face Transformers Trainer | Deep Learning

preview_player
Показать описание
Hugging Face provides three ways to fine-tune a pretrained text classification model: Tensorflow Keras, PyTorch, and transformer trainer. Transformer trainer is an API for feature-complete training in PyTorch without writing all the loops. This tutorial will use the transformer trainer to fine-tune a text classification model. We will talk about the following:

👉 How does transfer learning work?
👉 How to convert a pandas dataframe into a Hugging Face Dataset?
👉 How to tokenize text, load a pretrained model, set training arguments, and train a transfer learning model?
👉 How to make predictions and evaluate the model performance of a fine-tuned transfer learning model for text classification?
👉 How to save the model and re-load the model?

⏰ Timecodes ⏰
0:00 - Intro
00:52 - Step 0: Transfer Learning Algorithms
1:52 - Step 1: Install and Import Libraries
2:17 - Step 2: Download And Read Data
3:57 - Step 3: Train Test Split
4:25 - Step 4: Convert Pandas Dataframe to Hugging Face Dataset
5:41 - Step 5: Tokenize Text
9:13 - Step 6: Load Pretrained Model
9:45 - Step 7: Set Training Argument
11:42 - Step 8: Set Evaluation Metrics
12:10 - Step 9: Train Model Using Transformer Trainer
13:04 - Step 10: Make Predictions for Text Classification
14:40 - Step 11: Model Performance Evaluation
15:06 - Step 12: Save and Load Model

You will get full access to posts on Medium for $5 per month, and I will receive a portion of it. Thank you for your support!

📺 Videos mentioned in the video

🔥 Check out more machine learning tutorials on my website!

#transferlearning #finetuning #transformers #huggingface #grabngoinfo
Рекомендации по теме
Комментарии
Автор


You will get full access to posts on Medium for $5 per month, and I will receive a portion of it. Thank you for your support!

📺 Videos mentioned in the video

grabngoinfo
Автор

What does feature-complete training mean?

AnonymousIguana
Автор

Great work! Btw interested to know which text-to-speech software do you use?

sassydesi