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

convert-hf : fix duplicated block_count

Browse source
This commit is contained in:
Jared Van Bortel 2024-03-28 18:08:40 -04:00
parent 6dba2de027
commit 9bf29e6e8b
1 changed files with 0 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

1
convert-hf-to-gguf.py
View file

@ -1065,7 +1065,6 @@ class LlamaModel(Model):
def set_gguf_parameters(self): def set_gguf_parameters(self):
super().set_gguf_parameters() super().set_gguf_parameters()
hparams = self.hparams hparams = self.hparams
self.gguf_writer.add_block_count(hparams["num_hidden_layers"])
self.gguf_writer.add_rope_dimension_count(hparams["hidden_size"] // hparams["num_attention_heads"]) self.gguf_writer.add_rope_dimension_count(hparams["hidden_size"] // hparams["num_attention_heads"])
# Same as super class, but permuting q_proj, k_proj # Same as super class, but permuting q_proj, k_proj