Faster Text Generation with Self-Speculative Decoding
•
45
Mostafa Elhoushi
melhoushi
AI & ML interests
Make ML faster, smaller, and more accurate
Recent Activity
commented
a paper
20 days ago
LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding
updated
a model
about 2 months ago
facebook/layerskip-llama2-70B
updated
a collection
about 2 months ago
LayerSkip
Articles
Organizations
melhoushi's activity
No public activity