Update llama.py

This commit is contained in:
Daniel Han-Chen 2024-03-10 03:48:05 +11:00
commit daba749ee1

View file

@ -985,7 +985,9 @@ class FastLlamaModel:
token = token,
)
print(tokenizer)
model, tokenizer = patch_tokenizer(model, tokenizer)
print(tokenizer)
model = model_patcher.post_patch(model)
# Patch up QKV / O and MLP