tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-26 19:55:04 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama : add support for jina-reranker-v2 (#13900)

Browse Source
This commit is contained in:
Sigbjørn Skjæret
2025-05-29 21:42:31 +02:00
committed by GitHub
parent 2b131621e6
commit e83ba3e460
5 changed files with 119 additions and 72 deletions
Download Patch File Download Diff File Expand all files Collapse all files

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

@ -1036,6 +1036,7 @@ MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
MODEL_TENSOR.POS_EMBD,
MODEL_TENSOR.OUTPUT_NORM,
MODEL_TENSOR.ATTN_OUT_NORM,
MODEL_TENSOR.ATTN_QKV,
MODEL_TENSOR.ATTN_Q,
MODEL_TENSOR.ATTN_K,
MODEL_TENSOR.ATTN_V,

2
gguf-py/gguf/tensor_mapping.py
View File

@ -157,6 +157,7 @@ class TensorNameMap:
"h.{bid}.attn.c_attn", # gpt2
"transformer.h.{bid}.mixer.Wqkv", # phi2
"encoder.layers.{bid}.attn.Wqkv", # nomic-bert
"encoder.layers.{bid}.mixer.Wqkv", # jina
"model.layers.{bid}.self_attn.qkv_proj", # phi3
"encoder.layers.{bid}.self_attention.query_key_value", # chatglm
"transformer.layers.{bid}.attn.qkv_proj", # openelm
@ -224,6 +225,7 @@ class TensorNameMap:
"model.layers.layers.{bid}.self_attn.o_proj", # plamo
"model.layers.{bid}.attention.wo", # internlm2
"encoder.layers.{bid}.attn.out_proj", # nomic-bert
"encoder.layers.{bid}.mixer.out_proj", # jina
"transformer.decoder_layer.{bid}.multi_head_attention.linear", # Grok
"transformer.blocks.{bid}.norm_attn_norm.attn.out_proj", # dbrx
"encoder.layers.{bid}.self_attention.dense", # chatglm