conftrace_

multi-bucket data

1 paper

Explore in graph

Co-occurring keywords · papers mentioning both

data composition (1) fixed-length sequence (1) language model pretraining (1) padding ratio (1) truncation ratio (1) concatenation ratio (1) long-range dependency (1) training efficiency (1)

Papers

Beyond Fixed Length: Bucket Pre-training is All You Need IJCAI 2025