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

Update convert_hf_to_gguf.py

Browse source 
Co-authored-by: compilade <git@compilade.net>
This commit is contained in:
Jeffrey Morgan 2024-07-25 00:52:15 -04:00 • committed by jmorganca
parent e6bacb405a
commit 24540dd2bc
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

@ -1541,7 +1541,7 @@ class LlamaModel(Model):
rope_factors.append(1 / ((1 - smooth) / factor + smooth)) rope_factors.append(1 / ((1 - smooth) / factor + smooth))
self.gguf_writer.add_rope_scaling_attn_factors(1.0) self.gguf_writer.add_rope_scaling_attn_factors(1.0)
self.gguf_writer.add_tensor(gguf.TENSOR_NAMES[gguf.MODEL_TENSOR.ROPE_FREQS] + ".weight", np.array(rope_factors, dtype=np.float32)) self.gguf_writer.add_tensor(self.format_tensor_name(gguf.MODEL_TENSOR.ROPE_FREQS), np.array(rope_factors, dtype=np.float32))
@staticmethod @staticmethod
def permute(weights: Tensor, n_head: int, n_head_kv: int | None): def permute(weights: Tensor, n_head: int, n_head_kv: int | None):