valenti/spark-vibevoice
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix for windows

Browse Source
This commit is contained in:
mrfakename
2025-10-22 14:32:46 -07:00
committed by GitHub
parent 63a21e2b45
commit db4cad0723
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
vibevoice/modular/modeling_vibevoice.py
+2 -2
View File
@@ -259,9 +259,9 @@ class VibeVoiceForConditionalGeneration(VibeVoicePreTrainedModel):
"constant", "constant",
0, 0,
) )
print("✅ Tied input and output embeddings using standard assignment.") print("Tied input and output embeddings using standard assignment.")
else: else:
print("ℹ️ tie_word_embeddings is False, not tying weights.") print("tie_word_embeddings is False, not tying weights.")
# Also, ensure set_output_embeddings is safe, though your implementation looks okay. # Also, ensure set_output_embeddings is safe, though your implementation looks okay.
# The key is to avoid calling it after accelerator.prepare(). # The key is to avoid calling it after accelerator.prepare().