[SpecDecode] Add LiLiCorr drafter (#57934)

Signed-off-by: Andrii Skliar <askliar@nvidia.com>
Signed-off-by: Andrii Skliar <andreyws96@gmail.com>
Co-authored-by: Andrii Skliar <askliar@nvidia.com>
Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: Benjamin Chislett <chislett.ben@gmail.com>
This commit is contained in:
Andrii Skliar
2026-09-25 10:08:28 -04:00
committed by GitHub
co-authored by Andrii Skliar Codex Benjamin Chislett
parent 25b0add7b8
commit 73a78e6f1f
21 changed files with 1944 additions and 110 deletions
+7
View File
@@ -1502,6 +1502,13 @@ _SPECULATIVE_DECODING_EXAMPLE_MODELS = {
max_model_len=8192, # Reduce max len to ensure test runs in low-VRAM CI env
max_num_seqs=32,
),
# LiLiCorr checkpoints are not published yet.
"LiLiCorrDraftModel": _HfExamplesInfo(
"Qwen/Qwen3-8B",
speculative_model="LiLiCorrDraftModel",
is_available_online=False,
use_original_num_layers=True,
),
"DFlash2DraftModel": _HfExamplesInfo(
"Qwen/Qwen3.8-27B",
speculative_model="z-lab/Qwen3.8-27B-DFlash2",