The Glow-AI/QWEN2.5-14b_instruct-MSQ1000_step1-5_with_N_neg_batchsize256_best_epoch galaxy

7 models are built on Glow-AI/QWEN2.5-14b_instruct-MSQ1000_step1-5_with_N_neg_batchsize256_best_epoch: 0 quantized, 0 fine-tuned, 0 adapters and 7 merges. Together they were downloaded 93 times in the last 30 days.

Most downloaded direct derivatives

• yungisimon/Qwen2.5-14B-merge-task-sum-l0.8-nqa-msq-bcp (merge), 14 downloads in 30 days
• yungisimon/Qwen2.5-14B-merge-tiessum-r0.2-l1.0-nqa-msq-bcp (merge), 12 downloads in 30 days
• yungisimon/Qwen2.5-14B-merge-task-sum-l1.0-nqa-msq-bcp (merge), 10 downloads in 30 days
• yungisimon/Qwen2.5-14B-merge-ties-r0.2-nqa-msq-bcp (merge), 10 downloads in 30 days
• yungisimon/Qwen2.5-14B-merge-tiessum-r0.5-l0.7-nqa-msq-bcp (merge), 10 downloads in 30 days
• yungisimon/Qwen2.5-14B-ties-merge-nqa-msq-bcp (merge), 10 downloads in 30 days
• yungisimon/Qwen2.5-14B-merge-ties-r0.3-bcp2x-nqa-msq-bcp (merge), 9 downloads in 30 days

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