From daa97446beb934cfc1e2bf11133f5b103409478e Mon Sep 17 00:00:00 2001 From: "github-actions[bot]" <41898282+github-actions[bot]@users.noreply.github.com> Date: Fri, 21 Apr 2023 17:09:38 -0700 Subject: [PATCH] fix replace_bos_with_pad not found (#6443) (#6450) Signed-off-by: Abhinav Khattar Co-authored-by: Abhinav Khattar --- .../nlp/models/language_modeling/megatron_finetune_model.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/nemo/collections/nlp/models/language_modeling/megatron_finetune_model.py b/nemo/collections/nlp/models/language_modeling/megatron_finetune_model.py index 452819e1d5c4..7fc48856453f 100644 --- a/nemo/collections/nlp/models/language_modeling/megatron_finetune_model.py +++ b/nemo/collections/nlp/models/language_modeling/megatron_finetune_model.py @@ -331,7 +331,7 @@ def inference_step(self, dataloader_iter, batch_idx: int, mode: str, dataloader_ tokens_enc=batch['text_enc'], enc_mask=batch['enc_mask'], num_tokens_to_generate=30, - bos_id=self.tokenizer.pad_id if data_cfg.replace_bos_with_pad else self.tokenizer.bos_id, + bos_id=self.tokenizer.pad_id if data_cfg.get('replace_bos_with_pad', False) else self.tokenizer.bos_id, ) # Special ids to text function to handle stripping and special tokens with sentencepiece tokenizers.