dolphin

能否使用自己搭建的backbone网络训练，能否使用自己的数据集

1

我用自己搭建的mobilenet-v2跑训练，报错Graph disconnected: cannot obtain value for tensor Tensor("input_2:0", shape=(?, 224, 224, 3), dtype=float32) at layer "input_2". The following previous layers were accessed without issue: [] 另外我想训练自己的数据集，可以直接替换database文件，并按up主的格式整理数据，然后就可以训练了吗

super(AdamWeightDecayOptimizer, self).__init__(False, name) TypeError: __init__() takes 1 positional argument but 3 were give

2

Deprecated in favor of operator or tf.math.divide. INFO:tensorflow:Error recorded from training_loop: __init__() takes 1 positional argument but 3 were given INFO:tensorflow:training_loop marked as finished WARNING:tensorflow:Reraising captured error Traceback (most recent...

suffixes missing :S.A. DE C.V.

1

you miss a suffixes: S.A. DE C.V. please add it ,thanks

如上 Sent from PPHub

大家英文wordpiece都是怎么修改的呢，为什么我改过之后，预测出的标签有很多X，而token并没有那么多～预测出的标签远远大于输入字符串的wordpiece结果 Sent from PPHub

用了自己的英文NER数据集，删除了crf的tanh激活函数，开启服务就报错如上

服务启动后，服务端接收到数据但一直未返回结果，也未报错，请问是什么问题呢？

1

服务端一直处于这个状态 I:SINK:[__i:_ru:317]:ready I:WORKER-0:[__i:_ru:497]:use device cpu, load graph from /tf/train_out/classification_model.pb I:WORKER-0:[__i:gen:537]:ready and listening! I:VENTILATOR:[__i:_ru:215]:new encode request req id: 0 size: 3 client: b'7c94f7a6-2729-4ad4-b950-ff36f107b07e' I:SINK:[__i:_ru:369]:job register size: 3 job id: b'7c94f7a6-2729-4ad4-b950-ff36f107b07e#0' I:WORKER-0:[__i:gen:545]:new...

请问input_span_mask是什么？为什么要在Input Features中增加这个?没看懂

1

请问input_span_mask是什么？为什么要在Input Features中增加这个?没看懂

您好，calc_f1_score函数为什么返回的是max(f1_scores)呢？这部分可以解释一下吗，返回最大的，不是就是1了吗？这里没看明白

1

def calc_f1_score(answers, prediction): f1_scores = [] for ans in answers: ans_segs = mixed_segmentation(ans, rm_punc=True) prediction_segs = mixed_segmentation(prediction, rm_punc=True) lcs, lcs_len = find_lcs(ans_segs, prediction_segs) if ans == "" and prediction ==...

__new__() missing 2 required positional arguments: 'start_index' and 'end_index'

17

我修改该代码执行SQuAD2.0数据集，报错： WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/python/training/saver.py:1266: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version. Instructions for updating: Use standard file APIs to check for files with this prefix....