nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 21 Packages Projects Releases Wiki Activity

model : add EXAONE MoE (#18543)

Browse Source 
* Add EXAONE MoE implementations

Co-authored-by: Junwon Hwang <nuclear1221@gmail.com>

* Address PR feedback

* Address PR feedback

* [WIP] Add MTP for EXAONE-MoE

* Address PR feedback

* Address PR feedback

* Address PR feedback

* Address PR feedback

* Address PR feedback

* Address PR feedback

* Address PR feedback

---------

Co-authored-by: LG-AI-EXAONE <exaonemodels@lgresearch.ai>
This commit is contained in:
Junwon Hwang
2026-01-13 23:28:38 +01:00
committed by GitHub
co-authored by LG-AI-EXAONE
parent e4832e3ae4
commit 60591f01d4
17 changed files with 656 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
gguf-py/gguf/tensor_mapping.py
+3 -2
View File
@@ -436,7 +436,8 @@ class TensorNameMap:
"model.layers.{bid}.mlp.expert_bias", # afmoe
"model.layers.{bid}.feed_forward.expert_bias", # lfm2moe
"model.layers.{bid}.block_sparse_moe.e_score_correction", # minimax-m2
"backbone.layers.{bid}.mixer.gate.e_score_correction" # nemotron-h-moe
"backbone.layers.{bid}.mixer.gate.e_score_correction", # nemotron-h-moe
"model.layers.{bid}.mlp.e_score_correction", # exaone-moe
),
# Feed-forward up
@@ -1794,7 +1795,7 @@ class TensorNameMap:
"model.embed_audio.soft_embedding_norm", # gemma3n
),
# NextN/MTP tensors for GLM4_MOE
# NextN/MTP tensors
MODEL_TENSOR.NEXTN_EH_PROJ: (
"model.layers.{bid}.eh_proj",
),