nlpyang/hiersumm

Vocab file error

Opened this issue · 0 comments

Hi,

I got an error when reading the vocab file: spm9998_3.model
RuntimeError: Internal: /sentencepiece/python/bundled/sentencepiece/src/sentencepiece_processor.cc(823) [model_proto->ParseFromArray(serialized.data(), serialized.size())]

When I google it, they said the file is not in the correct format.
Could you please help ASAP

here is the train instruction:
python train_abstractive.py -data_path data/ranked_wiki_b40/WIKI -mode train -batch_size 10500 -seed 666 -train_steps 1000000 -save_checkpoint_steps 5000 -report_every 100 -trunc_tgt_ntoken 400 -trunc_src_nblock 24 -visible_gpus 0 -gpu_ranks 0 -world_size 1 -accum_count 1 -dec_dropout 0.1 -enc_dropout 0.1 -label_smoothing 0.1 -vocab_path data/spm9998_3.model -model_path model/ -log_file log/log -inter_layers 6,7 -inter_heads 8 -hier