tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-27 03:55:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

convert : fix tensor naming conflict for llama 4 vision (#13836)

Browse Source 
* convert : fix tensor naming conflict for llama 4 vision

* add comment
This commit is contained in:
Xuan-Son Nguyen
2025-05-28 10:05:54 +02:00
committed by GitHub
parent 1e8659e65a
commit 26b79b6cb3
2 changed files with 3 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
convert_hf_to_gguf.py
View File

@ -2169,6 +2169,9 @@ class Llama4VisionModel(MmprojModel):
# process vision tensors # process vision tensors
if "positional_embedding_vlm" in name and ".weight" not in name: if "positional_embedding_vlm" in name and ".weight" not in name:
name += ".weight" name += ".weight"
if "multi_modal_projector.linear_1" in name:
# despite the name with number postfix, this is a single fully connected layer
return [(gguf.TENSOR_NAMES[gguf.MODEL_TENSOR.V_MMPROJ_FC], data_torch)]
return [(self.map_tensor_name(name), data_torch)] return [(self.map_tensor_name(name), data_torch)]
return [] return []

1
gguf-py/gguf/tensor_mapping.py
View File

@ -902,7 +902,6 @@ class TensorNameMap:
MODEL_TENSOR.V_MMPROJ_FC: ( MODEL_TENSOR.V_MMPROJ_FC: (
"model.connector.modality_projection.proj", # SmolVLM "model.connector.modality_projection.proj", # SmolVLM
"multi_modal_projector.linear_1", # llama 4
), ),
MODEL_TENSOR.V_MMPROJ_MLP: ( MODEL_TENSOR.V_MMPROJ_MLP: (