t2ance/LCB-Selection-Data-8192

Name: LCB Selection Data 8192 (VERL format)
Creator: t2ance

LCB Selection Data 8192 (VERL format): RL dataset on Hugging Face. VERL-compatible parquet flatten of LCB selection data, used as the data source for current GRPO training runs (4B / 8B / bs96-nrp variants) on both the legacy bilevel-reweighting training stack and the VERL training stack.

t2ance/LCB-Selection-Data-8192 on the Hugging Face Hub