novarobot
/
llama-cpu
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update merge-weights.py

Browse Source
main
randaller 3 years ago committed by GitHub
parent 485c8dbba3
commit 0238fb92ee
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 0 additions and 3 deletions
Show Stats Download Patch File Download Diff File

3
merge-weights.py
Unescape Escape View File

@ -60,7 +60,6 @@ def write_model(input_base_path, model_size):
for layer_i in range(n_layers): for layer_i in range(n_layers):
if model_size == "7B": if model_size == "7B":
# Unsharded
state_dict |= { state_dict |= {
f"layers.{layer_i}.attention.wq.weight": loaded[ f"layers.{layer_i}.attention.wq.weight": loaded[
f"layers.{layer_i}.attention.wq.weight" f"layers.{layer_i}.attention.wq.weight"
@ -89,7 +88,6 @@ def write_model(input_base_path, model_size):
f"layers.{layer_i}.ffn_norm.weight": loaded[f"layers.{layer_i}.ffn_norm.weight"], f"layers.{layer_i}.ffn_norm.weight": loaded[f"layers.{layer_i}.ffn_norm.weight"],
} }
else: else:
# Sharded
state_dict |= { state_dict |= {
f"layers.{layer_i}.attention_norm.weight": loaded[0][ f"layers.{layer_i}.attention_norm.weight": loaded[0][
f"layers.{layer_i}.attention_norm.weight" f"layers.{layer_i}.attention_norm.weight"
@ -131,7 +129,6 @@ def write_model(input_base_path, model_size):
) )
if model_size == "7B": if model_size == "7B":
# Unsharded
state_dict |= { state_dict |= {
"tok_embeddings.weight": loaded["tok_embeddings.weight"], "tok_embeddings.weight": loaded["tok_embeddings.weight"],
"norm.weight": loaded["norm.weight"], "norm.weight": loaded["norm.weight"],

Write Preview
Loading…
Cancel
Save