Studying the impacts of pre-training using ChatGPT-generated text on downstream tasks.
Sarthak AnandPublished in: CoRR (2023)
Keyphrases
- database
- automatically extracted
- training process
- text retrieval
- training algorithm
- automatically generated
- data sets
- human generated
- textual data
- test set
- training examples
- online learning
- text mining
- information retrieval
- training samples
- free text
- multi task
- supply chain
- keywords
- training phase
- natural language generation
- manually constructed
- data mining
- text fragments
- natural language processing systems