Scaling Up Models and Data with t5x and seqio.
Adam RobertsHyung Won ChungAnselm LevskayaGaurav MishraJames BradburyDaniel AndorSharan NarangBrian LesterColin GaffneyAfroz MohiuddinCurtis HawthorneAitor LewkowyczAlex SalcianuMarc van ZeeJacob AustinSebastian GoodmanLivio Baldini SoaresHaitang HuSasha TsvyashchenkoAakanksha ChowdheryJasmijn BastingsJannis BulianXavier GarciaJianmo NiAndrew ChenKathleen KenealyJonathan H. ClarkStephan LeeDan GarretteJames Lee-ThorpColin RaffelNoam ShazeerMarvin RitterMaarten BosmaAlexandre PassosJeremy Maitin-ShepardNoah FiedelMark OmernickBrennan SaetaRyan SepassiAlexander SpiridonovJoshua NewlanAndrea GesmundoPublished in: CoRR (2022)
Keyphrases
- data sets
- historical data
- statistical analysis
- prior knowledge
- data sources
- image data
- data objects
- training data
- high dimensional data
- data structure
- data processing
- subject specific
- empirical data
- raw data
- end users
- data analysis
- database
- learned models
- data collection
- neural network
- sensor data
- machine learning
- learning models
- knowledge discovery
- statistical methods
- missing values
- incomplete data
- classification models
- feature selection
- original data
- website
- data distribution
- experimental data
- spatial data
- high quality
- data mining techniques
- probability distribution