Ellie Cheng
@ellieyhc.bsky.social
PhD student @mit_csail | Programming Systems for AI/ML | don’t ask how many cups of coffee I’ve drank
When it’s cold I’m eepy when it’s hot I’m eepy. The eepydemic is real
April 24, 2025 at 7:25 PM
When it’s cold I’m eepy when it’s hot I’m eepy. The eepydemic is real
Mentally I’m still at breakfast drinking my coffee
April 17, 2025 at 4:41 PM
Mentally I’m still at breakfast drinking my coffee
Reposted by Ellie Cheng
Excited to share our work with friends from MIT/Google on Learned Asynchronous Decoding! LLM responses often contain chunks of tokens that are semantically independent. What if we can train LLMs to identify such chunks and decode them in parallel, thereby speeding up inference? 1/N
February 27, 2025 at 12:38 AM
Excited to share our work with friends from MIT/Google on Learned Asynchronous Decoding! LLM responses often contain chunks of tokens that are semantically independent. What if we can train LLMs to identify such chunks and decode them in parallel, thereby speeding up inference? 1/N