vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Address comments by @compilade

Browse source
This commit is contained in:
Yoshi Suhara 2024-08-09 02:11:43 -07:00
parent 092382fee3
commit 6f369f3ffa
1 changed files with 0 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

2
convert_hf_to_gguf.py
View file

@ -3765,8 +3765,6 @@ class NemotronModel(Model):
self.gguf_writer.add_rope_scaling_type(gguf.RopeScalingType.LINEAR)
self.gguf_writer.add_rope_scaling_factor(self.hparams["factor"])
_experts: list[dict[str, Tensor]] | None = None
def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iterable[tuple[str, Tensor]]:
# * Adding +1 to LayerNorm's weights here to implement layernorm1p w/o changing anything on the GGML engine side
# model.layers.{l}.input_layernorm.weight