Official Inference Code of EAGLE-4
docs: lead the README with adaptive, always-on speculation
Seyfal Sultanovf71e245a45ac
diff --git a/README.md b/README.mdindex 1fc0224..8d6408a 100644--- a/README.md+++ b/README.md@@ -49,15 +49,16 @@## EAGLE-4-EAGLE-4 (Extrapolation Algorithm for Greater Language-model Efficiency) speeds up LLM inference with speculative-decoding. One draft model adapts its draft tree to the batch size, the hardware and the workload, and keeps the-target model's output distribution. On four B300 GPUs it serves Qwen3.8-27B 3.42× faster than SGLang without-speculation for one request at a time.+**Leave speculation on. EAGLE-4 adapts to your hardware, batch and workload.**-This repository is `eagle-harness`, the package that runs EAGLE-4 drafts inside SGLang, vLLM and TokenSpeed.+EAGLE-4 (Extrapolation Algorithm for Greater Language-model Efficiency) speeds up LLM inference with+speculative decoding. EAGLE-4 draft models are built to run with `eagle-harness`, which decides how far to+speculate from measured costs on your setup, so every draft tree fits the batch size, the hardware and the+workload. The output stays exactly what the target model would generate on its own: up to 3.4× faster+generation on Qwen3.8-27B, with the same answers. [Measured speedups](#measured-speedup)-**Note that an EAGLE-4 draft model is only one component of EAGLE-4 and does not represent the complete system-on its own. It must be combined with EAGLE-Harness to form the full EAGLE-4 system.**+This repository holds `eagle-harness`, the package that brings EAGLE-4 to SGLang, vLLM and TokenSpeed.+Install it, point your engine at an EAGLE-4 draft model, and it serves with EAGLE-4.## Quick start