Segmentation fault error
#2
by
yangxia20000
- opened
I got Segmentation fault when I ran vocab_transplant.py. Besides, tokenizer_target.vocab_size is only 128000, not 128256, why do we use tokenizer_target.vocab_size rather than target_vocab_size for model.lm_head.out_features ? Thanks a lot!
yangxia20000
changed discussion status to
closed