dl-nlp-using-paddlenlp

paddlenlp打卡笔记，待整理。所有的学习材料可在基于深度学习的自然语言处理课程上找到（已结课），只需在Paddle AI Studio注册一个账号即可。paddlenlp文件夹内储存的是课程学习笔记和项目练习，但涉及到模型训练的部分，绝大大多数需要在GPU上运行，所以推荐直接在Paddle AI Studio平台上注册课程、直接运行。paddlenlp_updated_notes_English文件夹是我会陆续更新的新的、更为系统化的学习心得，并且也更容易看懂，可以直接在 Google Colab或者你的电脑上运行。两个版本的笔记是不一样的。英语的笔记不是中文笔记的直接翻译。

Learning notes and materials on paddlenlp, taken during and after the Deep Learning Based Natural Language Processing (The course is over, but you can still sign up for the course materials). The paddlenlp folder stores the course materials as well as my notes taken during the course (in Chinese), but most of the notebooks regarding model trainning need to be run on Paddle AI Studio platform for free GPU resources. The paddlenlp_updated_notes_English folder is a more systematic and beginner-friendly collection of notes that I will update regularly in English and can be run directly on Google Colab or your own computers. Therefore, the contents of the two folders are very different. The English notes are not the translations to the Chinese ones.

目录 Table of contents

(The links are for the updated English notes stored in paddlenlp_updated_notes_English folder that have been updated.)

• 1-词向量 Word Embedding
  • loading pre-trained word embedding in paddlenp.ipynb
  • calculating text cosine similarity in paddlenlp.ipynb
  • embedding visualization using paddlepaddle tool.ipynb
  • embedding visualization using tensorflow tool.ipynb
  • training word embeddings using skip-gram with negative sampling in paddle.ipynb

• 2-文本语义相似度计算 Text similarity
  • Quick starts with paddle.ipynb
  • Get data.ipynb
  • preprocess data.ipynb
  • wrapped up data preprocessor.ipynb
  • preprocess data using paddle.ipynb

• 3-NER 命名实体识别 Named Entity Recognition

• 4-情感分析 Sentiment analysis
  • Quick starts with paddle.ipynb
  • Get data.ipynb
  • preprocess data.ipynb
  • wrapped up data preprocessor.ipynb
  • preprocess data using paddle.ipynb

• 5-实体关系抽取 Relation extraction
• 6-问答系统/机器阅读理解/段落检索 Q&A system/Machine Reading Comprehension/passage retrieval
• 7-结构化数据问答 Structured Q&A system
• 8-文本翻译 Machine Translation
• 9-机器同传 Machine Interpretation
• 10-任务式对话 Task-Oriented Dialogue System
• 11-开放域对话 Open-domain Dialogue System
• 12-产业实践 Industry practice (pretraining)