TokenButler -- Predict token importance for all heads across the transformer in the first layer itself. Enable fine-grained token sparsity!
YASH AKHAURI
akhauriyash
AI & ML interests
None yet
Recent Activity
updated
a model
about 20 hours ago
akhauriyash/E2EGRPO_bm14B_32Gen_8GAcc_2K_2xAccF
upvoted
a
paper
2 days ago
Performance Prediction for Large Systems via Text-to-Text Regression
published
a model
9 days ago
akhauriyash/E2EGRPO_bm14B_32Gen_8GAcc_2K_2xAccF
Organizations
None yet
models
37
akhauriyash/E2EGRPO_bm14B_32Gen_8GAcc_2K_2xAccF
2B
•
Updated
•
22
akhauriyash/E2EGRPO_bm14B_36Gen_2K_2xAccF
Updated
akhauriyash/E2EGRPO_bm14B_128Gen_2K_2xAccF
Updated
akhauriyash/E2EGRPO_bm14B_128Gen_2K_2xAcc
Updated
akhauriyash/E2EGRPO_bm14B_64Gen_4K_2xAcc
Updated
akhauriyash/DeepSeek-R1-Distill-Qwen-1.5B-E2EGRPO-OpenR1_Math_SpecR_GRPO_Mini-MiniSet_14BDrafter
2B
•
Updated
•
8
akhauriyash/DeepSeek-R1-Distill-Qwen-1.5B-E2EGRPO-OpenR1_Math_SpecR_GRPO_Mini-MiniSet
2B
•
Updated
•
52
akhauriyash/DeepSeek-R1-Distill-Qwen-1.5B-E2EGRPO-OpenR1_Math_SpecR_GRPO_Mini-MiniSet_32BDrafter
Updated
akhauriyash/DeepSeek-R1-Distill-Qwen-1.5B-E2EGRPO-OpenR1-220K
2B
•
Updated
•
8
akhauriyash/DeepSeek-R1-Distill-Qwen-1.5B-GRPO-SpeculativeReasoner_Mini
Text Generation
•
2B
•
Updated
•
27