hi-todayis-jh/l0-qwen3-1.7b-compression-bs32-n16-32k-t1-no-eos-seqmean-verl091-146102-rollouts download history

hi-todayis-jh/l0-qwen3-1.7b-compression-bs32-n16-32k-t1-no-eos-seqmean-verl091-146102-rollouts is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 458 times (458 in the last 7 days), and 458 times in total. It ranks #35,612 among datasets by monthly downloads.

RL training rollouts l0_Qwen3-1.7B_compression_bs32_n16_32k_t1_no_eos_verl091_seqmean One verified gzip JSONL shard per training step; 512 responses per shard. Historical compression MathVerify after thinking, without an EOS gate.

Open hi-todayis-jh/l0-qwen3-1.7b-compression-bs32-n16-32k-t1-no-eos-seqmean-verl091-146102-rollouts on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.