Datasets collection and preprocessings framework for NLP extreme multitask learning
nlp benchmark text-classification glue scaling natural-language-inference crossfit multi-task-learning meta-learning bigbench discriminative huggingface dataset-collection instruction-tuning preprocessings multi-task-learning-scaling extreme-multi-task-learning extreme-mtl curated-datasets reward-modeling
-
Updated
Jul 10, 2024 - Python