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

Fix hard coded layer_norm_eps

Browse source
This commit is contained in:
Anas Ahouzi 2024-02-24 12:46:35 -08:00
parent 8a529990fa
commit d4a952e099
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

@ -1077,7 +1077,7 @@ class StableLMModel(Model):
self.gguf_writer.add_rope_dimension_count(int(hparams["partial_rotary_factor"] * (hparams["hidden_size"] // hparams["num_attention_heads"])))
self.gguf_writer.add_head_count(hparams["num_attention_heads"])
self.gguf_writer.add_parallel_residual(hparams["use_parallel_residual"] if "use_parallel_residual" in hparams else True)
self.gguf_writer.add_layer_norm_eps(1e-5)
self.gguf_writer.add_layer_norm_eps(hparams["layer_norm_eps"])
class MixtralModel(Model):