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

Missing variable

Browse source
This commit is contained in:
Galunid 2023-10-29 02:06:41 +01:00
parent 989db34149
commit 550b925af2
1 changed files with 6 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

6
model.py
View file

@ -427,6 +427,12 @@ class BaichuanModel(Model):
head_count = self.hparams["num_attention_heads"] head_count = self.hparams["num_attention_heads"]
tensor_map = gguf.get_tensor_name_map(self.model_arch, block_count) tensor_map = gguf.get_tensor_name_map(self.model_arch, block_count)
if "num_key_value_heads" in self.hparams:
head_count_kv = self.hparams["num_key_value_heads"]
else:
head_count_kv = head_count
for i in range(block_count): for i in range(block_count):
if f"model.layers.{i}.self_attn.W_pack.weight" in model_kv: if f"model.layers.{i}.self_attn.W_pack.weight" in model_kv:
print(f"Unpacking and permuting layer {i}") print(f"Unpacking and permuting layer {i}")