minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

fix(eagle3): drop +1 offset on aux layer ids when first id != 1 (#25454)

Browse Source
This commit is contained in:
Qiaolin Yu
2026-05-18 11:25:51 -07:00
committed by GitHub
parent 86c6c77f2f
commit f5049709b3
1 changed files with 6 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/sglang/srt/models/deepseek_v2.py
+6 -3
View File
@@ -2564,9 +2564,12 @@ class DeepseekV2ForCausalLM(nn.Module, DeepseekV2WeightLoaderMixin):
self.model.layers_to_capture = [2, num_layers // 2, num_layers - 3] self.model.layers_to_capture = [2, num_layers // 2, num_layers - 3]
else: else:
self.capture_aux_hidden_states = True self.capture_aux_hidden_states = True
# we plus 1 here because in sglang, for the ith layer, it takes the output # TODO (Qiaolin-Yu): check if other draft models need similar layer id
# of the (i-1)th layer as aux hidden state # adjustment
self.model.layers_to_capture = [val + 1 for val in layer_ids] if layer_ids and layer_ids[0] == 1:
self.model.layers_to_capture = [val + 1 for val in layer_ids]
else:
self.model.layers_to_capture = list(layer_ids)
def set_dflash_layers_to_capture(self, layer_ids: List[int]): def set_dflash_layers_to_capture(self, layer_ids: List[int]):
if not self.pp_group.is_last_rank: if not self.pp_group.is_last_rank: