Fix(tokenizer): Fix condition to add pad token (#477)

* Fix(tokenizer): Fix condition to add pad token

* chore: fix lint
This commit is contained in:
NanoCode012
2023-08-25 14:30:50 +09:00
committed by GitHub
parent cb9797ef5a
commit 71bd06243c

View File

@@ -55,10 +55,15 @@ def load_tokenizer(cfg):
**tokenizer_kwargs,
)
if tokenizer.__class__.__name__ in [
"LlamaTokenizer",
"LlamaTokenizerFast",
] and not hasattr(tokenizer, "pad_token"):
if (
tokenizer.__class__.__name__
in [
"LlamaTokenizer",
"LlamaTokenizerFast",
]
and hasattr(tokenizer, "pad_token")
and not tokenizer.pad_token
):
# set a pad_token, but use eos_token so we don't add a new token
tokenizer.pad_token = LLAMA_DEFAULT_EOS_TOKEN