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

convert-*.py: need to include self in per_model_weight_count_estimation()

Browse source
This commit is contained in:
brian khuu 2024-06-04 02:18:53 +10:00
parent 54918ad14e
commit 39472a09da
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
convert_hf_to_gguf.py
View file

@ -426,7 +426,7 @@ class Model:
self.gguf_writer.write_kv_data_to_file()
self.gguf_writer.close()
def per_model_weight_count_estimation(tensors: Iterator[tuple[str, Tensor]], expert_count: int) -> int:
def per_model_weight_count_estimation(self, tensors: Iterator[tuple[str, Tensor]], expert_count: int) -> int:
# TODO: Ensure parameter count is accurate throughout various model type
# May currently overestimate parameter count in Mamba model because
# output weights is tied with token embeddings.