关于tokenizer的改动:
/mnt/nas/pretrain/code/Megatron-LM/megatron/tokenizer/__init__.py 或者 tokenizer.py
在build_tokenizer.py函数中:
?
elif args.tokenizer_type == "QwenTokenizer":
assert args.tokenizer_name_or_path is not None
from .tokenization_qwen import QWenTokenizer
tokenizer = QWenTokenizer.from_pretrained(
args.tokenizer_name_or_path,
model_max_length=args.seq_length,
padding_side='right',
use_fast=False,
)
tokenizer.pad_token_id = tokenizer.pad_id
tokenizer.eos_token_id = tokenizer.eod_id
args.padded_vocab_size = tokenizer.vocab_size + args.extra_vocab_size
?
debug时