Model checkpoints for paper "A Token is Worth over 1,000 Tokens: Efficient Knowledge Distillation through Low-Rank Clone".
Jitai Hao
JitaiHao
AI & ML interests
None yet
Organizations
models 13
JitaiHao/Qwen2.5-7B-Instruct-1M-Compressor
0.2B • Updated • 11
JitaiHao/Qwen2.5-32B-Instruct-Compressor
1B • Updated • 7
JitaiHao/Llama-3.1-8B-Instruct-Compressor
0.5B • Updated • 6
JitaiHao/Uni-X-3B
4B • Updated • 11
JitaiHao/Uni-7B
Text Generation • 8B • Updated • 4 • 1
JitaiHao/Shuo-v0.1-SFT
Updated
JitaiHao/Shuo-v0.1-4B
Updated
JitaiHao/LRC-4B-SFT
4B • Updated • 6
JitaiHao/LRC-4B-Base
4B • Updated • 7
JitaiHao/LRC-1.7B-SFT
2B • Updated • 3
datasets 5
JitaiHao/deltakv_qwen_train_vr1.0_num40000_seqlen8192
Preview • Updated • 27
JitaiHao/deltakv_qwen_train_v3.0_num40000_seqlen8192
Viewer • Updated • 33k • 64
JitaiHao/deltakv_qwen_train_num80000_seqlen8192
Viewer • Updated • 80k • 89
JitaiHao/deltakv_qwen_train_num10000seqlen4096
Viewer • Updated • 60k • 593
JitaiHao/deltakv_llama3_train_num40000_seqlen8192
Viewer • Updated • 40k • 78