A new Speculative decoding drafter for vLLM and SGLang called EAGLE-4
EAGLE-4 is a new speculative-decoding drafter for vLLM and SGLang that may speed up inference by producing draft tokens more efficiently, though real-world gain
Original sourceEAGLE-4 is a new speculative-decoding drafter for vLLM and SGLang that may speed up inference by producing draft tokens more efficiently, though real-world gain
Original source