LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention
Paper
•
2502.14866
•
Published
•
13
Shang Yang
Shangy
AI & ML interests
None yet
Recent Activity
authored
a paper
about 1 month ago
AWQ: Activation-aware Weight Quantization for LLM Compression and
Acceleration
authored
a paper
about 1 month ago
QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM
Serving
authored
a paper
about 1 month ago
FlatFormer: Flattened Window Attention for Efficient Point Cloud
Transformer
Organizations
