CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-27 08:21:30 +00:00
Code Issues Packages Projects Releases Wiki Activity

split q_proj/gate

Browse Source
This commit is contained in:
Xuan Son Nguyen
2025-09-20 12:00:14 +07:00
parent 178230ee21
commit 46110e0630
6 changed files with 31 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
gguf-py/gguf/constants.py
View File

@@ -433,6 +433,7 @@ class MODEL_TENSOR(IntEnum):
ATTN_NORM_2 = auto()
ATTN_OUT_NORM = auto()
ATTN_POST_NORM = auto()
ATTN_GATE = auto()
ATTN_ROT_EMBD = auto()
ATTN_SINKS = auto()
FFN_GATE_INP = auto()
@@ -776,6 +777,7 @@ TENSOR_NAMES: dict[MODEL_TENSOR, str] = {
MODEL_TENSOR.ATTN_K_NORM: "blk.{bid}.attn_k_norm",
MODEL_TENSOR.ATTN_OUT_NORM: "blk.{bid}.attn_output_norm",
MODEL_TENSOR.ATTN_POST_NORM: "blk.{bid}.post_attention_norm",
MODEL_TENSOR.ATTN_GATE: "blk.{bid}.attn_gate",
MODEL_TENSOR.FFN_GATE_INP: "blk.{bid}.ffn_gate_inp",
MODEL_TENSOR.FFN_GATE_INP_SHEXP: "blk.{bid}.ffn_gate_inp_shexp",
MODEL_TENSOR.FFN_NORM: "blk.{bid}.ffn_norm",
@@ -1478,6 +1480,7 @@ MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
MODEL_TENSOR.ATTN_V,
MODEL_TENSOR.ATTN_OUT,
MODEL_TENSOR.ATTN_POST_NORM,
MODEL_TENSOR.ATTN_GATE,
MODEL_TENSOR.FFN_GATE_INP,
MODEL_TENSOR.FFN_GATE_INP_SHEXP,
MODEL_TENSOR.FFN_UP_SHEXP,