diff options
Diffstat (limited to 'gguf-py')
| -rw-r--r-- | gguf-py/gguf/constants.py | 1 | ||||
| -rw-r--r-- | gguf-py/gguf/tensor_mapping.py | 3 |
2 files changed, 4 insertions, 0 deletions
diff --git a/gguf-py/gguf/constants.py b/gguf-py/gguf/constants.py index a3c024c8..8908585c 100644 --- a/gguf-py/gguf/constants.py +++ b/gguf-py/gguf/constants.py @@ -415,6 +415,7 @@ MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = { MODEL_TENSOR.TOKEN_EMBD, MODEL_TENSOR.TOKEN_EMBD_NORM, MODEL_TENSOR.TOKEN_TYPES, + MODEL_TENSOR.ATTN_NORM_2, MODEL_TENSOR.ATTN_OUT_NORM, MODEL_TENSOR.ATTN_Q, MODEL_TENSOR.ATTN_Q_NORM, diff --git a/gguf-py/gguf/tensor_mapping.py b/gguf-py/gguf/tensor_mapping.py index 83e3c4c3..81b4992a 100644 --- a/gguf-py/gguf/tensor_mapping.py +++ b/gguf-py/gguf/tensor_mapping.py @@ -102,6 +102,7 @@ class TensorNameMap: # Attention norm 2 MODEL_TENSOR.ATTN_NORM_2: ( "transformer.h.{bid}.ln_attn", # falcon40b + "encoder.layer.{bid}.layer_norm_1", # jina-v2-code ), # Attention query-key-value @@ -311,6 +312,7 @@ class TensorNameMap: "model.layers.{bid}.mlp.c_proj", # starcoder2 "encoder.layer.{bid}.mlp.wo", # jina-bert-v2 "model.layers.{bid}.residual_mlp.w2", # arctic + "encoder.layer.{bid}.mlp.down_layer", # jina-bert-v2 ), MODEL_TENSOR.FFN_DOWN_EXP: ( @@ -350,6 +352,7 @@ class TensorNameMap: "encoder.layers.{bid}.norm2", # nomic-bert "transformer.decoder_layer.{bid}.rms_norm_3", # Grok "encoder.layer.{bid}.mlp.layernorm", # jina-bert-v2 + "encoder.layer.{bid}.layer_norm_2" # jina-v2-code ), MODEL_TENSOR.SSM_IN: ( |