我一直在嘗試使用Pegasus 庫生成摘要并按照上述步驟操作 -創(chuàng)建輸入數(shù)據(jù) .tfrecord于pegasus\data\testdatatransformer_params創(chuàng)建一個返回命名的函數(shù)test_transformers(假設(shè))跑步 python3 pegasus/bin/train.py --params=test_transformer --param_overrides=vocab_filename=ckpt/pegasus_ckpt/c4.unigram.newline.10pct.96000.model,batch_size=1,beam_size=5,beam_alpha=0.6 --model_dir=ckpt/pegasus_ckpt/xsum/model.ckpt-30000python3 pegasus/bin/evaluate.py --params=test_transformer --param_overrides=vocab_filename=ckpt/pegasus_ckpt/c4.unigram.newline.10pct.96000.model,batch_size=1,beam_size=5,beam_alpha=0.6 --model_dir=ckpt/pegasus_ckpt/xsum/model.ckpt-30000但是,當我生成文本時,我在輸出中遇到這個問題 -它的實現(xiàn)方式或我在步驟 3 和 4 中運行 python 代碼的方式是否存在問題?
1 回答

牧羊人nacy
TA貢獻1862條經(jīng)驗 獲得超7個贊
此問題突出的原因是:-
1. --model_dir is typically a directory instead of a particular checkpoint.?
? ?-> Try changing model_dir to actual model directory instead of checkpoint
2. It seems there are only 100 training steps.?
? ?-> Try changing "train_steps": 100
添加回答
舉報
0/150
提交
取消