biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Bugfix] Fix Positive Feature Layers in Llava Models (#13514)

Browse Source 
Signed-off-by: Alex-Brooks <Alex.brooks@ibm.com>
This commit is contained in:
Alex Brooks
2025-02-19 01:50:07 -07:00
committed by GitHub
parent fdc5df6f54
commit 983a40a8bb
6 changed files with 44 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
vllm/model_executor/models/llava.py
View File

@@ -428,7 +428,7 @@ def _get_num_hidden_layers(hf_config: LlavaLikeConfig) -> int:
def _get_layer_index(feature_layer_index: int, num_hidden_layers: int) -> int:
"""Given an signed vision feature layer, get the number of hidden layers
"""Given a signed vision feature layer, get the number of hidden layers
needed to leverage it.
Args:
@@ -438,7 +438,7 @@ def _get_layer_index(feature_layer_index: int, num_hidden_layers: int) -> int:
"""
if feature_layer_index < 0:
return num_hidden_layers + feature_layer_index + 1
return feature_layer_index + 1
return feature_layer_index
def init_vision_tower_for_llava(