A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 8, 10:00 UTC
[Docs] Spec decoding docs warning removal (#34439)
Signed-off-by: NickLucche <nlucches@redhat.com>
Nicolò Lucchesi committed
Feb 12, 2026 at 18:01 UTC
334c715e0f4f4de2d3de90bd0b9bba59df143eda
1 file changed
-5
docs/features/spec_decode/README.md
-5
@@ -1,10 +1,5 @@
1
# Speculative Decoding
2
3
-!!! warning
4
- Please note that speculative decoding in vLLM is not yet optimized and does
5
- not usually yield inter-token latency reductions for all prompt datasets or sampling parameters.
6
- The work to optimize it is ongoing and can be followed here: <https://github.com/vllm-project/vllm/issues/4630>
7
-
3
!!! warning
4
Currently, speculative decoding in vLLM is not compatible with pipeline parallelism.
5