Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-01-26 03:12:23 +01:00
Code Issues Packages Projects Releases Wiki Activity

Remove .attention from skipped tensors to match more accurately (#7051)

Browse Source
This commit is contained in:
Bartowski 2024-05-02 19:49:09 -04:00 committed by GitHub
parent 6ecf3189e0
commit 60325fa56f
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
convert-hf-to-gguf.py
View File

@ -1427,7 +1427,7 @@ class LlamaModel(Model):
experts = dict() experts = dict()
for name, data_torch in self.get_tensors(): for name, data_torch in self.get_tensors():
# we don't need these # we don't need these
if name.endswith((".attention.masked_bias", ".attention.bias", ".attention.rotary_emb.inv_freq")): if name.endswith((".attention.masked_bias", ".attention.bias", ".rotary_emb.inv_freq")):
continue continue
old_dtype = data_torch.dtype old_dtype = data_torch.dtype