llama : support for Llama-3_1-Nemotron-51B (#10669)

* conflict resolution * move comments after bracket to its own line
2024-12-23 08:22:33 +08:00 · 2024-12-23 08:22:33 +08:00 · 6f0c9e034b
commit 6f0c9e034b
parent dab76c92cc
4 changed files with 471 additions and 1 deletions
--- a/gguf-py/gguf/tensor_mapping.py
+++ b/gguf-py/gguf/tensor_mapping.py
@ -198,6 +198,7 @@ class TensorNameMap:
            "transformer.h.{bid}.self_attention.dense",                     # falcon
            "h.{bid}.self_attention.dense",                                 # bloom
            "model.layers.{bid}.self_attn.o_proj",                          # llama-hf nemotron olmoe olmo2
+            "model.layers.{bid}.self_attn.linear_attn",                     # deci
            "layers.{bid}.attention.wo",                                    # llama-pth
            "encoder.layer.{bid}.attention.output.dense",                   # bert
            "transformer.h.{bid}.attn.out_proj",                            # gpt-j