cnn_encoder RuntimeError

[PetrochukM]

Input: {"sentence": "what is i miss you"}

Trace:

Traceback (most recent call last):
  File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/usr/lib/python3.6/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/run.py", line 13, in <module>
    main(prog="python -m allennlp.run")
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/__init__.py", line 77, in main
    args.func(args)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/predict.py", line 145, in predict_inner
    _run(predictor, input_file, output_file, args.batch_size, not args.silent, args.cuda_device)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/predict.py", line 120, in _run
    _run_predictor(batch_json_data)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/predict.py", line 98, in _run_predictor
    result = predictor.predict_json(batch_data[0], cuda_device)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/service/predictors/sentence_tagger.py", line 40, in predict_json
    output = self._model.forward_on_instance(instance, cuda_device)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/models/model.py", line 117, in forward_on_instance
    outputs = self.decode(self.forward(**model_input))
  File "/home/michael/Desktop/lattice/allennlp/allennlp/models/simple_tagger.py", line 99, in forward
    embedded_text_input = self.text_field_embedder(tokens)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/text_field_embedders/basic_text_field_embedder.py", line 47, in forward
    token_vectors = embedder(tensor)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/token_embedders/token_characters_encoder.py", line 36, in forward
    return self._dropout(self._encoder(self._embedding(token_characters), mask))
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/time_distributed.py", line 35, in forward
    reshaped_outputs = self._module(*reshaped_inputs)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/seq2vec_encoders/cnn_encoder.py", line 103, in forward
    for convolution_layer in self._convolution_layers]
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/seq2vec_encoders/cnn_encoder.py", line 103, in <listcomp>
    for convolution_layer in self._convolution_layers]
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/conv.py", line 154, in forward
    self.padding, self.dilation, self.groups)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/functional.py", line 83, in conv1d
    return f(input, weight, bias)
RuntimeError: Given input size: (8 x 1 x 4). Calculated output size: (50 x 1 x 0). Output size is too small at /pytorch/torch/lib/THNN/generic/SpatialConvolutionMM.c:45

Full Output:

2017-12-04 13:42:43,598 - INFO - allennlp.models.archival - extracting archive file /tmp/subject_recognition/model.tar.gz to temp dir /tmp/tmpidl36u2j
2017-12-04 13:42:43,649 - INFO - allennlp.data.vocabulary - Loading token dictionary from /tmp/tmpidl36u2j/vocabulary.
2017-12-04 13:42:43,658 - INFO - allennlp.common.params - model.type = simple_tagger
2017-12-04 13:42:43,658 - INFO - allennlp.common.params - model.text_field_embedder.type = basic
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.type = embedding
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.num_embeddings = None
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.vocab_namespace = tokens
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.embedding_dim = 50
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.pretrained_file = None
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.projection_dim = None
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.trainable = True
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.padding_index = None
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.max_norm = None
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.norm_type = 2.0
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.scale_grad_by_freq = False
2017-12-04 13:42:43,659 - INFO - allennlp.common.params - model.text_field_embedder.tokens.sparse = False
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.type = character_encoding
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.num_embeddings = None
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.vocab_namespace = token_characters
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.embedding_dim = 8
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.pretrained_file = None
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.projection_dim = None
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.trainable = True
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.padding_index = None
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.max_norm = None
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.norm_type = 2.0
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.scale_grad_by_freq = False
2017-12-04 13:42:43,693 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.embedding.sparse = False
2017-12-04 13:42:43,694 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.encoder.type = cnn
2017-12-04 13:42:43,694 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.encoder.embedding_dim = 8
2017-12-04 13:42:43,694 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.encoder.output_dim = None
2017-12-04 13:42:43,694 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.encoder.num_filters = 50
2017-12-04 13:42:43,694 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.encoder.conv_layer_activation = relu
2017-12-04 13:42:43,694 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.encoder.ngram_filter_sizes = [5]
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.text_field_embedder.token_characters.dropout = 0.2
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.stacked_encoder.type = lstm
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.stacked_encoder.batch_first = True
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - Converting Params object to dict; logging of default values will not occur when dictionary parameters are used subsequently.
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - CURRENTLY DEFINED PARAMETERS:
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.stacked_encoder.input_size = 100
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.stacked_encoder.hidden_size = 100
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.stacked_encoder.num_layers = 2
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.stacked_encoder.dropout = 0.5
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.stacked_encoder.bidirectional = True
2017-12-04 13:42:43,695 - INFO - allennlp.common.params - model.stacked_encoder.batch_first = True
2017-12-04 13:42:43,698 - INFO - allennlp.common.params - model.initializer = []
2017-12-04 13:42:43,698 - INFO - allennlp.common.params - model.regularizer = []
2017-12-04 13:42:43,698 - INFO - allennlp.nn.initializers - Initializing parameters
2017-12-04 13:42:43,698 - INFO - allennlp.nn.initializers - Done initializing parameters; the following parameters are using their default initialization from their code
2017-12-04 13:42:43,698 - INFO - allennlp.nn.initializers -    stacked_encoder._module.bias_hh_l0
2017-12-04 13:42:43,698 - INFO - allennlp.nn.initializers -    stacked_encoder._module.bias_hh_l0_reverse
2017-12-04 13:42:43,698 - INFO - allennlp.nn.initializers -    stacked_encoder._module.bias_hh_l1
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.bias_hh_l1_reverse
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.bias_ih_l0
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.bias_ih_l0_reverse
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.bias_ih_l1
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.bias_ih_l1_reverse
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.weight_hh_l0
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.weight_hh_l0_reverse
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.weight_hh_l1
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.weight_hh_l1_reverse
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.weight_ih_l0
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.weight_ih_l0_reverse
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.weight_ih_l1
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    stacked_encoder._module.weight_ih_l1_reverse
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    tag_projection_layer._module.bias
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    tag_projection_layer._module.weight
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    text_field_embedder.token_embedder_token_characters._embedding._module.weight
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    text_field_embedder.token_embedder_token_characters._encoder._module.conv_layer_0.bias
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    text_field_embedder.token_embedder_token_characters._encoder._module.conv_layer_0.weight
2017-12-04 13:42:43,699 - INFO - allennlp.nn.initializers -    text_field_embedder.token_embedder_tokens.weight
2017-12-04 13:42:43,703 - INFO - allennlp.common.params - dataset_reader.type = sequence_tagging
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.tokens.type = single_id
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.tokens.namespace = tokens
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.tokens.lowercase_tokens = True
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.token_characters.type = characters
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.token_characters.namespace = token_characters
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.token_characters.character_tokenizer.byte_encoding = None
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.token_characters.character_tokenizer.lowercase_characters = False
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.token_characters.character_tokenizer.start_tokens = None
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_indexers.token_characters.character_tokenizer.end_tokens = None
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.word_tag_delimiter = /
2017-12-04 13:42:43,704 - INFO - allennlp.common.params - dataset_reader.token_delimiter = None
input:  {'sentence': 'who was the trump ocean club international hotel and tower named after'}
prediction:  {"logits": [[6.398567199707031, -6.493162155151367], [7.395605087280273, -7.58927583694458], [5.744316101074219, -6.178823471069336], [-4.630782127380371, 4.574059009552002], [-7.62890100479126, 7.8646345138549805], [-6.244271755218506,6.426706314086914], [-6.991970062255859, 7.288210868835449], [-7.308270454406738, 7.667025566101074], [-7.963583946228027, 8.456028938293457], [-7.350199222564697, 7.405206680297852], [7.853301525115967, -8.249849319458008], [5.914466857910156, -6.401645660400391]], "class_probabilities": [[0.9999974966049194, 2.5187900973833166e-06], [0.9999997019767761, 3.1056248417371535e-07], [0.9999933838844299, 6.635037607338745e-06], [0.00010054128506453708, 0.9998994469642639], [1.8674231228033022e-07, 0.9999998211860657], [3.1409613256982993e-06, 0.9999968409538269], [6.283418088059989e-07, 0.9999993443489075], [3.135533575004956e-07, 0.9999997019767761], [7.396950252314127e-08, 0.9999999403953552], [3.906687595645053e-07, 0.9999995827674866], [0.9999998807907104, 1.0150564833111275e-07], [0.9999955296516418, 4.478972186916508e-06]], "tags": ["O", "O", "O", "I", "I", "I", "I", "I", "I", "I", "O", "O"], "words": ["who", "was", "the", "trump", "ocean", "club", "international", "hotel", "and", "tower", "named", "after"]}
input:  {'sentence': 'where was sasha vujačić born'}
prediction:  {"logits": [[8.635701179504395, -8.88974380493164], [9.944202423095703, -10.215315818786621], [-9.7186918258667, 9.900626182556152], [-9.876214027404785, 9.75756549835205], [7.849674701690674, -8.177754402160645]], "class_probabilities": [[1.0, 2.447915115055821e-08], [1.0, 1.7572439059421185e-09], [3.0160494102204893e-09, 1.0], [2.9727467154572196e-09, 1.0], [0.9999998807907104, 1.0949031548079802e-07]], "tags": ["O", "O", "I", "I", "O"], "words": ["where", "was", "sasha", "vuja\u010di\u0107", "born"]}
input:  {'sentence': 'what is a region that dead combo was released in'}
prediction:  {"logits": [[8.803048133850098, -9.027806282043457], [8.533225059509277, -8.732645988464355], [8.513354301452637, -8.759966850280762], [7.975431442260742, -8.300684928894043], [7.844444274902344, -8.313096046447754], [-6.992236137390137, 6.945400238037109], [-4.740862846374512, 4.410414218902588], [7.65218448638916, -8.091584205627441], [7.788219451904297, -8.322957038879395], [7.016716957092285, -7.5209197998046875]], "class_probabilities": [[1.0, 1.803675075962019e-08], [0.9999999403953552, 3.1734195715671376e-08], [0.9999999403953552, 3.14986507987669e-08], [0.9999999403953552, 8.538328444274157e-08], [0.9999998807907104, 9.613239626560244e-08], [8.850362291923375e-07, 0.9999991059303284], [0.00010607293370412663, 0.9998939037322998], [0.9999998807907104, 1.4540123061124177e-07], [0.9999998807907104, 1.0069440747884073e-07], [0.9999995231628418, 4.857182034356811e-07]], "tags": ["O", "O", "O", "O", "O", "I", "I", "O", "O", "O"], "words": ["what", "is", "a", "region", "that", "dead", "combo", "was", "released", "in"]}
input:  {'sentence': 'what is a film directed by wiebke von carolsfeld'}
prediction:  {"logits": [[8.623019218444824, -8.82130241394043], [8.561206817626953, -8.759057998657227], [8.64100170135498, -8.875571250915527], [8.759867668151855, -9.020319938659668], [9.330001831054688, -9.71264362335205], [8.674726486206055, -8.978828430175781], [-8.949294090270996, 8.996564865112305], [-8.993559837341309, 9.218438148498535], [-6.377027988433838, 6.554698944091797]], "class_probabilities": [[1.0, 2.654775421717659e-08], [0.9999999403953552, 3.0054156496817086e-08], [1.0, 2.4697273559581845e-08], [1.0, 1.8974162685481133e-08], [1.0, 5.368890665380377e-09], [1.0, 2.1535676353323652e-08], [1.607729238628508e-08, 1.0], [1.2320541031840548e-08, 1.0], [2.4200342068070313e-06, 0.9999975562095642]], "tags": ["O", "O", "O", "O", "O", "O", "I", "I", "I"], "words": ["what", "is", "a", "film", "directed", "by", "wiebke", "von", "carolsfeld"]}
input:  {'sentence': 'what country was music for stock exchange released in'}
prediction:  {"logits": [[7.261748313903809, -7.322234153747559], [6.869565010070801, -7.116580009460449], [7.612556457519531, -7.895277976989746], [-8.300959587097168, 8.384926795959473], [-7.1197357177734375, 7.408495903015137], [-7.4351701736450195, 7.880918502807617], [-6.292874813079834, 6.31099796295166], [7.562020778656006, -8.098215103149414], [6.345880508422852, -6.879611968994141]], "class_probabilities": [[0.9999995231628418, 4.6372090878321615e-07], [0.9999991655349731, 8.431289870713954e-07], [0.9999998211860657, 1.84091192068081e-07], [5.6677617266132074e-08, 0.9999999403953552], [4.903079684481781e-07, 0.9999995231628418], [2.2300116597762099e-07, 0.9999997615814209], [3.358971753186779e-06, 0.9999966621398926], [0.9999998211860657, 1.5806863018497097e-07], [0.9999982118606567, 1.8040177565126214e-06]], "tags": ["O", "O", "O", "I", "I", "I", "I", "O", "O"], "words": ["what", "country", "was", "music", "for", "stock", "exchange", "released", "in"]}
input:  {'sentence': 'where is adler school of professional psychology located'}
prediction:  {"logits": [[8.046847343444824, -8.120144844055176], [8.013388633728027, -8.230134963989258], [-6.277460098266602, 6.4742584228515625], [-5.907905578613281, 6.11325740814209], [-5.706315040588379, 6.040461540222168], [-5.762022018432617, 6.161492347717285], [-4.986045837402344, 4.969719886779785], [8.554655075073242, -8.924757957458496]], "class_probabilities": [[0.9999998807907104, 9.522796062810812e-08], [0.9999999403953552, 8.821201191722139e-08], [2.8973286134714726e-06, 0.999997079372406], [6.015512553858571e-06, 0.9999939799308777], [7.914733942016028e-06, 0.9999920725822449], [6.632551503571449e-06, 0.9999933838844299], [4.7450990678044036e-05, 0.9999525547027588], [1.0, 2.563231227270535e-08]], "tags": ["O", "O", "I", "I", "I", "I", "I", "O"], "words": ["where", "is", "adler", "school", "of", "professional", "psychology", "located"]}
input:  {'sentence': 'where was john miltern born'}
prediction:  {"logits": [[9.069466590881348, -9.35698127746582], [9.840865135192871, -10.077906608581543], [-10.508658409118652, 10.750466346740723], [-9.011463165283203, 9.049300193786621], [6.570859909057617, -6.856064319610596]], "class_probabilities": [[1.0, 9.942485057479189e-09], [1.0, 2.2355652884442634e-09], [5.851664708345083e-10, 1.0], [1.4332096398561589e-08, 1.0], [0.9999985098838806, 1.4748928833796526e-06]], "tags": ["O", "O", "I", "I", "O"], "words": ["where", "was", "john", "miltern", "born"]}
input:  {'sentence': 'what city is vancouver millionaires from'}
prediction:  {"logits": [[9.478272438049316, -9.723587989807129], [8.991626739501953, -9.41037368774414], [8.994111061096191, -9.270081520080566], [-9.780631065368652, 10.112456321716309], [-8.470414161682129, 8.292500495910645], [8.871359825134277,-9.238316535949707]], "class_probabilities": [[1.0, 4.578655676823473e-09], [1.0, 1.0188558441370787e-08], [1.0, 1.1693969348414157e-08], [2.293728096347536e-09, 1.0], [5.247574819122747e-08, 0.9999999403953552], [1.0, 1.3647952101791816e-08]], "tags": ["O", "O", "O", "I", "I", "O"], "words": ["what", "city", "is", "vancouver", "millionaires", "from"]}
input:  {'sentence': "what was seymour parker gilbert's profession"}
prediction:  {"logits": [[7.263981819152832, -7.445630073547363], [7.406196594238281, -7.779941082000732], [-3.2452712059020996, 3.2185041904449463], [-4.820518970489502, 5.070255756378174], [-5.884641170501709, 6.167430400848389], [-2.3690390586853027, 2.381502628326416], [6.158773422241211, -6.47823429107666]], "class_probabilities": [[0.9999995827674866, 4.089747847046965e-07], [0.9999997615814209, 2.539474905915995e-07], [0.0015564723871648312, 0.9984435439109802], [5.063713251729496e-05, 0.9999493360519409], [5.8324267229181714e-06, 0.999994158744812], [0.008572880178689957, 0.9914271235466003], [0.9999967217445374, 3.2494951938133454e-06]], "tags": ["O", "O", "I", "I", "I", "I", "O"], "words": ["what", "was", "seymour", "parker", "gilbert", "'s", "profession"]}
input:  {'sentence': 'what does ( 12385 ) 1994 uo orbit'}
prediction:  {"logits": [[8.661669731140137, -8.73204231262207], [6.756413459777832, -6.913588523864746], [-8.12545108795166, 8.37150764465332], [-6.21560525894165, 6.574915885925293], [-5.908961296081543, 6.351534843444824], [-6.748265266418457, 7.286136627197266], [-6.559242248535156, 6.733588218688965], [5.455531597137451, -5.718820571899414]], "class_probabilities": [[1.0, 2.7925905143888485e-08], [0.9999988675117493, 1.1566262401174754e-06], [6.846386213510414e-08, 0.9999999403953552], [2.787056928355014e-06, 0.9999971985816956], [4.735132733912906e-06, 0.9999952912330627], [8.034083975871908e-07, 0.9999992251396179], [1.6865388943188009e-06, 0.9999983310699463], [0.9999859929084778, 1.4029242265678477e-05]], "tags": ["O", "O", "I", "I", "I", "I", "I", "O"], "words": ["what", "does", "(", "12385", ")", "1994", "uo", "orbit"]}
input:  {'sentence': 'who is the singer of only women bleed'}
prediction:  {"logits": [[8.146283149719238, -8.417633056640625], [8.382765769958496, -8.649081230163574], [8.707167625427246, -9.06165599822998], [5.899953842163086, -6.245193958282471], [6.676791191101074, -7.133207321166992], [-7.3403215408325195, 7.326386451721191], [-8.70136833190918, 9.017411231994629], [-7.2209320068359375, 7.519026279449463]], "class_probabilities": [[0.9999999403953552, 6.402991914455924e-08], [0.9999999403953552, 4.010170329138418e-08], [1.0, 1.9191014999364597e-08], [0.999994695186615, 5.314069312589709e-06], [0.9999989867210388, 1.0055262009700527e-06], [4.269032558568142e-07, 0.9999995827674866], [2.0175829007484936e-08, 1.0], [3.96750493791842e-07, 0.9999995827674866]], "tags": ["O", "O", "O", "O", "O", "I", "I", "I"], "words": ["who", "is", "the", "singer", "of", "only", "women", "bleed"]}
input:  {'sentence': 'in what french city did antoine de févin die'}
prediction:  {"logits": [[8.42149829864502, -8.597514152526855], [9.159464836120605, -9.398354530334473], [7.911324501037598, -8.196645736694336], [8.41940689086914, -8.864227294921875], [8.473825454711914, -8.763897895812988], [-7.265119552612305, 7.268061637878418], [-9.695326805114746, 10.123551368713379], [-8.6390962600708, 8.721709251403809], [6.088113784790039, -6.572582244873047]], "class_probabilities": [[0.9999999403953552, 4.061971026203537e-08], [1.0, 8.718493482717804e-09], [0.9999998807907104, 1.010175836313465e-07], [0.9999999403953552, 3.117547109354746e-08], [0.9999999403953552, 3.264009862391504e-08], [4.878871777691529e-07, 0.9999995231628418], [2.470418758449e-09, 1.0], [2.886010896929747e-08, 1.0], [0.9999968409538269, 3.1734248295833822e-06]], "tags": ["O", "O", "O", "O", "O", "I", "I", "I", "O"], "words": ["in", "what", "french", "city", "did", "antoine", "de", "f\u00e9vin", "die"]}
input:  {'sentence': 'who published rama'}
prediction:  {"logits": [[7.790890216827393, -8.061838150024414], [8.480996131896973, -8.880188941955566], [-8.509429931640625, 8.673094749450684]], "class_probabilities": [[0.9999998807907104, 1.3039102952916437e-07], [1.0, 2.884915595302573e-08], [3.449244090347747e-08, 0.9999999403953552]], "tags": ["O", "O", "I"], "words": ["who", "published", "rama"]}
input:  {'sentence': 'who was an advisor for irving langmuir'}
prediction:  {"logits": [[7.9313764572143555, -8.197521209716797], [11.116253852844238, -11.250279426574707], [7.747143268585205, -7.930570125579834], [8.567495346069336, -8.810744285583496], [8.90336799621582, -9.314750671386719], [-7.9317827224731445, 7.93057918548584], [-6.167613983154297, 6.267350196838379]], "class_probabilities": [[0.9999998807907104, 9.892551133816596e-08], [1.0, 1.933472548731885e-10], [0.9999998211860657, 1.553300563728044e-07], [1.0, 2.8361345272287508e-08], [1.0, 1.2245361169505031e-08], [1.2914087221815862e-07, 0.9999998807907104], [3.977057986048749e-06, 0.9999960064888]], "tags": ["O", "O", "O", "O", "O", "I", "I"], "words": ["who", "was", "an", "advisor", "for", "irving", "langmuir"]}
input:  {'sentence': 'what is the language of the film bon voyage'}
prediction:  {"logits": [[9.125393867492676, -9.433004379272461], [8.793717384338379, -9.153564453125], [9.641524314880371, -10.056965827941895], [8.722613334655762, -9.140792846679688], [8.96627140045166, -9.304070472717285], [8.557679176330566, -9.08573055267334], [8.263012886047363, -8.70583724975586], [-9.588225364685059, 9.644758224487305], [-6.113242149353027, 6.316921234130859]], "class_probabilities": [[1.0, 8.71343974750971e-09], [1.0, 1.605443245011884e-08], [1.0, 2.786470387761142e-09], [1.0, 1.7459059975521996e-08], [1.0, 1.1622280027268062e-08], [1.0, 2.175527136216715e-08], [0.9999999403953552, 4.2709295655640744e-08], [4.438352352309494e-09, 1.0], [3.996196937805507e-06, 0.9999960064888]], "tags": ["O", "O", "O", "O", "O", "O", "O", "I", "I"], "words": ["what", "is", "the", "language", "of", "the", "film", "bon", "voyage"]}
input:  {'sentence': 'which country was the hunyadi family from'}
prediction:  {"logits": [[7.6912641525268555, -7.988090515136719], [7.4113287925720215, -7.695474624633789], [7.103708267211914, -7.419437408447266], [5.362257957458496, -5.7446184158325195], [-7.060808181762695, 7.138975620269775], [-6.7796525955200195, 6.628928184509277], [7.922821998596191, -8.39511775970459]], "class_probabilities": [[0.9999998211860657, 1.550753268020344e-07], [0.9999997019767761, 2.749148961811443e-07], [0.9999995231628418, 4.928079988530953e-07], [0.9999849796295166, 1.500853704783367e-05], [6.809452202105604e-07, 0.9999993443489075], [1.5021963690742268e-06, 0.9999985098838806], [0.9999999403953552, 8.188582967250113e-08]], "tags": ["O", "O", "O", "O", "I", "I", "O"], "words": ["which", "country", "was", "the", "hunyadi", "family", "from"]}
input:  {'sentence': 'what major cities does u . s . route 2 run through'}
prediction:  {"logits": [[8.969206809997559, -9.133158683776855], [7.795598983764648, -7.96260929107666], [8.794605255126953, -9.131536483764648], [9.340428352355957, -9.687677383422852], [-7.811139106750488, 7.866705417633057], [-7.043241500854492,7.295038223266602], [-6.032283782958984, 6.342902183532715], [-7.4129638671875, 7.7712812423706055], [-8.264636993408203, 8.8204984664917], [-7.390346527099609, 7.517102241516113], [4.612483978271484, -5.0306806564331055], [6.656200408935547, -6.983220100402832]], "class_probabilities": [[1.0, 1.3748095994969844e-08], [0.9999998807907104, 1.4331678244161594e-07], [1.0, 1.6397422086811275e-08], [1.0, 5.447512663181442e-09], [1.5530962116372393e-07, 0.9999998211860657], [5.928761765972013e-07, 0.9999994039535522], [4.2220481191179715e-06, 0.999995768070221], [2.5442867013225623e-07, 0.9999997615814209], [3.802064796332161e-08, 0.9999999403953552], [3.3556534617673606e-07, 0.9999996423721313], [0.9999351501464844, 6.48632412776351e-05], [0.9999988079071045, 1.1925440048798919e-06]], "tags": ["O", "O", "O", "O", "I", "I", "I", "I", "I", "I", "O", "O"], "words": ["what", "major", "cities", "does", "u", ".", "s", ".", "route", "2", "run", "through"]}
input:  {'sentence': 'who was a child of mithibai jinnah'}
prediction:  {"logits": [[8.075671195983887, -8.302603721618652], [10.268364906311035, -10.371853828430176], [7.979487419128418, -8.166990280151367], [7.125223159790039, -7.347713947296143], [7.909624099731445, -8.288236618041992], [-6.892998218536377, 6.918696880340576], [-6.853404998779297, 7.103806495666504]], "class_probabilities": [[0.9999999403953552, 7.709132887612213e-08], [1.0, 1.0865930377690347e-09], [0.9999998807907104, 9.72017772937761e-08], [0.999999463558197, 5.181831284062355e-07], [0.9999998807907104, 9.233331610403184e-08], [1.0038216942120926e-06, 0.9999989867210388], [8.678800327288627e-07, 0.9999991059303284]], "tags": ["O", "O", "O", "O", "O", "I", "I"], "words": ["who", "was", "a", "child", "of", "mithibai", "jinnah"]}
input:  {'sentence': 'whats a version of the single titled star'}
prediction:  {"logits": [[8.669629096984863, -9.015080451965332], [8.101340293884277, -8.336298942565918], [8.043974876403809, -8.295292854309082], [8.587550163269043, -8.855189323425293], [9.097281455993652, -9.411031723022461], [9.3046236038208, -9.746444702148438], [8.331998825073242, -8.765419006347656], [8.3409423828125, -8.758827209472656], [-8.345796585083008, 8.25795841217041]], "class_probabilities": [[1.0, 2.0875082995530647e-08], [0.9999999403953552, 7.264804935402935e-08], [0.9999999403953552, 8.01578678988335e-08], [1.0, 2.6589765056428405e-08], [1.0, 9.160983971412406e-09], [1.0, 5.323859575412371e-09], [0.9999999403953552, 3.75565569754599e-08], [0.9999999403953552, 3.7468335989387924e-08], [6.152907161549592e-08, 0.9999999403953552]], "tags": ["O", "O", "O", "O", "O", "O", "O", "O", "I"], "words": ["what", "s", "a", "version", "of", "the", "single", "titled", "star"]}
input:  {'sentence': 'what is a song by john rutter'}
prediction:  {"logits": [[8.387832641601562, -8.72761058807373], [8.51089096069336, -8.667384147644043], [8.829692840576172, -9.086806297302246], [9.462985038757324, -9.826141357421875], [8.844443321228027, -9.097152709960938], [-9.277384757995605, 9.561005592346191], [-5.950255870819092, 6.284684181213379]], "class_probabilities": [[0.9999999403953552, 3.688561633907739e-08], [0.9999999403953552, 3.463933140324116e-08], [1.0, 1.655628345531568e-08], [1.0, 4.196028413616659e-09], [1.0, 1.6145943249057382e-08], [6.585534340075583e-09, 1.0], [4.857704880123492e-06, 0.9999951720237732]], "tags": ["O", "O", "O", "O", "O", "I", "I"], "words": ["what", "is", "a", "song", "by", "john", "rutter"]}
input:  {'sentence': 'what job does jamie hewlett have'}
prediction:  {"logits": [[9.688754081726074, -9.971457481384277], [8.863424301147461, -9.303208351135254], [9.23021411895752, -9.529818534851074], [-10.300615310668945, 10.592827796936035], [-8.897722244262695, 8.828034400939941], [8.114100456237793, -8.5026273727417]], "class_probabilities": [[1.0, 2.8952003017224115e-09], [1.0, 1.2892325429447737e-08], [1.0, 7.122317846608439e-09], [8.435144804863626e-10, 1.0], [2.0035550107877498e-08, 1.0], [0.9999999403953552, 6.073607750067822e-08]], "tags": ["O", "O", "O", "I", "I", "O"], "words": ["what", "job", "does", "jamie", "hewlett", "have"]}
input:  {'sentence': "what's an example of an album"}
prediction:  {"logits": [[8.61514663696289, -8.673218727111816], [7.004127502441406, -7.123215675354004], [8.224970817565918, -8.324880599975586], [7.339296340942383, -7.378694534301758], [7.709934234619141, -7.961277008056641], [5.687994956970215, -5.9427995681762695], [-4.362086772918701, 4.18479061126709]], "class_probabilities": [[0.9999999403953552, 3.102835322010833e-08], [0.9999992847442627, 7.321035582208424e-07], [0.9999999403953552, 6.493684878705608e-08], [0.9999995827674866, 4.055623037402256e-07], [0.9999998211860657, 1.5634333294656244e-07], [0.9999911189079285, 8.888045158528257e-06], [0.00019411281391512603, 0.9998058676719666]], "tags": ["O", "O", "O", "O", "O", "O", "I"], "words": ["what", "'s", "an", "example", "of", "an", "album"]}
input:  {'sentence': 'which recordings contains the composition australia'}
prediction:  {"logits": [[8.819635391235352, -9.076542854309082], [6.902050971984863, -7.162337303161621], [8.849668502807617, -9.063918113708496], [9.761978149414062, -10.227490425109863], [7.922916412353516, -8.36593246459961], [-9.610908508300781, 9.746048927307129]], "class_probabilities": [[1.0, 1.6896164467539165e-08], [0.9999992251396179, 7.796746785970754e-07], [1.0, 1.6604607466774723e-08], [1.0, 2.0829771241182016e-09], [0.9999999403953552, 8.430295395100984e-08], [3.920853419714376e-09, 1.0]], "tags": ["O", "O", "O", "O", "O", "I"], "words": ["which", "recordings", "contains", "the", "composition", "australia"]}
input:  {'sentence': 'what is the film tempo di uccidere about'}
prediction:  {"logits": [[8.598097801208496, -8.817706108093262], [7.734391689300537, -8.0142240524292], [8.181905746459961, -8.672520637512207], [7.259640693664551, -7.715563774108887], [-7.542301177978516, 7.543377876281738], [-9.037955284118652, 9.440078735351562], [-6.657838344573975, 6.925506114959717], [6.128909111022949, -6.297570705413818]], "class_probabilities": [[1.0, 2.7315708805986105e-08], [0.9999998807907104, 1.4469823383933544e-07], [0.9999999403953552, 4.788672924860293e-08], [0.9999997019767761, 3.1358206342702033e-07], [2.807842349739076e-07, 0.9999997019767761], [9.442595150233046e-09, 1.0], [1.261327611246088e-06, 0.9999987483024597], [0.9999960064888, 4.01094666813151e-06]], "tags": ["O", "O", "O", "O", "I", "I", "I","O"], "words": ["what", "is", "the", "film", "tempo", "di", "uccidere", "about"]}
input:  {'sentence': 'what country is ghost house from'}
prediction:  {"logits": [[9.651347160339355, -9.90041446685791], [9.53153133392334, -9.98112964630127], [9.936897277832031, -10.30695629119873], [-9.088891983032227, 9.191853523254395], [-8.778704643249512, 8.79553508758545], [5.692631721496582, -6.184737205505371]], "class_probabilities": [[1.0, 3.2268427929693644e-09], [1.0, 3.3555136447205314e-09], [1.0, 1.6151233683814326e-09], [1.1501982477568617e-08, 1.0], [2.3313349473141898e-08, 1.0], [0.9999930262565613, 6.945782843104098e-06]], "tags": ["O", "O", "O", "I", "I", "O"], "words": ["what", "country", "is", "ghost", "house", "from"]}
input:  {'sentence': 'which country was the yamakinkarudu movie produced'}
prediction:  {"logits": [[6.902523994445801, -7.137324810028076], [6.246228218078613, -6.433320045471191], [6.852069854736328, -7.1025590896606445], [5.290034294128418, -5.628162384033203], [-6.307599067687988, 6.199684143066406], [4.734820365905762, -4.7280168533325195], [7.943081855773926, -8.113024711608887]], "class_probabilities": [[0.9999992251396179, 7.990445851646655e-07], [0.9999969005584717, 3.114158971584402e-06], [0.9999991059303284, 8.701242677489063e-07], [0.9999818801879883, 1.8125063434126787e-05], [3.699596163642127e-06, 0.9999963045120239], [0.9999223351478577, 7.767983333906159e-05], [0.9999998807907104, 1.0639505632070723e-07]], "tags": ["O", "O", "O", "O", "I", "O", "O"], "words": ["which", "country", "was", "the", "yamakinkarudu", "movie", "produced"]}
input:  {'sentence': "what's the time zone in sub-saharan africa"}
prediction:  {"logits": [[6.356539726257324, -6.414700984954834], [5.796345233917236, -5.9224348068237305], [5.36168098449707, -5.672167778015137], [3.2818074226379395, -3.396571636199951], [3.2155704498291016, -3.3407294750213623], [3.0466079711914062, -3.205622434616089], [-5.163661956787109, 5.1666059494018555], [-7.320122718811035, 7.556535720825195], [-6.925727844238281, 7.220663547515869], [-6.064815521240234, 6.200611591339111]], "class_probabilities": [[0.9999971389770508, 2.8413162453944096e-06], [0.9999918341636658, 8.139443707477767e-06], [0.9999838471412659, 1.614557004359085e-05], [0.9987437725067139, 0.0012562350602820516], [0.9985808730125427, 0.001419117208570242], [0.9980775713920593, 0.0019224497955292463], [3.2629279303364456e-05, 0.9999673962593079], [3.460582433945092e-07, 0.9999996423721313], [7.18290607437666e-07, 0.9999992847442627], [4.711843303084606e-06, 0.9999952912330627]], "tags": ["O", "O", "O", "O", "O", "O", "I", "I", "I", "I"], "words": ["what", "'s","the", "time", "zone", "in", "sub", "-", "saharan", "africa"]}
input:  {'sentence': 'what author wrote the book liquor'}
prediction:  {"logits": [[8.466856956481934, -8.7299222946167], [8.413829803466797, -8.68161678314209], [9.587078094482422, -9.88249683380127], [9.11201000213623, -9.526382446289062], [8.729208946228027, -9.1183500289917], [-8.674453735351562, 8.752230644226074]], "class_probabilities": [[0.9999999403953552, 3.400428738586925e-08], [0.9999999403953552, 3.763062750294921e-08], [1.0, 3.5032525769196354e-09], [1.0, 8.04356492523084e-09], [1.0, 1.773795865744887e-08], [2.7020085724416276e-08, 1.0]], "tags": ["O", "O", "O", "O", "O", "I"], "words": ["what", "author", "wrote", "the", "book", "liquor"]}
input:  {'sentence': 'what is the release type of the album wake'}
prediction:  {"logits": [[8.158788681030273, -8.61927318572998], [7.844760417938232, -8.110114097595215], [9.074728965759277, -9.461857795715332], [8.513115882873535, -8.90260124206543], [8.998818397521973, -9.410029411315918], [9.718024253845215, -10.148905754089355], [6.653080940246582, -7.043368816375732], [7.999353885650635, -8.313262939453125], [-8.426223754882812, 8.570143699645996]], "class_probabilities": [[0.9999999403953552, 5.1686924962268677e-08], [0.9999998807907104, 1.1772974062296271e-07], [1.0, 8.905589155006055e-09], [1.0, 2.7318053597014114e-08], [1.0, 1.0119031834676662e-08], [1.0, 2.3545174698824667e-09], [0.9999988675117493, 1.1264377235420397e-06], [0.9999999403953552, 8.232290582554924e-08], [4.154999544425664e-08, 0.9999999403953552]], "tags": ["O", "O", "O", "O", "O", "O", "O", "O", "I"], "words": ["what", "is", "the", "release", "type", "of", "the", "album", "wake"]}
input:  {'sentence': 'who is the chid of fritz leiber'}
prediction:  {"logits": [[8.853079795837402, -9.129118919372559], [11.219265937805176, -11.360259056091309], [6.700904846191406, -6.947626113891602], [5.472049236297607, -5.712995529174805], [6.917994499206543, -7.464934349060059], [-6.873827934265137, 6.839973449707031], [-6.667637825012207, 6.892522811889648]], "class_probabilities": [[1.0, 1.5503520245374602e-08], [1.0, 1.5625606364366718e-10], [0.9999988079071045, 1.181728748633759e-06], [0.9999861121177673, 1.3880034202884417e-05], [0.9999994039535522, 5.669870120073028e-07], [1.1070602567997412e-06, 0.9999988675117493], [1.2909115412185201e-06, 0.9999986886978149]], "tags": ["O", "O", "O", "O", "O", "I", "I"], "words": ["who", "is", "the", "chid", "of", "fritz", "leiber"]}
input:  {'sentence': 'what artist creates riot grrrl music'}
prediction:  {"logits": [[8.635003089904785, -8.896241188049316], [9.282937049865723, -9.65794849395752], [8.207551002502441, -8.428279876708984], [-4.560591697692871, 4.5571088790893555], [-6.988924980163574, 7.181877136230469], [6.050742149353027,-6.409419059753418]], "class_probabilities": [[1.0, 2.433757728681485e-08], [1.0, 5.9439879684930474e-09], [0.9999999403953552, 5.9586906075992374e-08], [0.00010969461436616257, 0.9998903274536133], [7.009685418779554e-07, 0.9999992847442627], [0.9999961256980896, 3.878100415022345e-06]], "tags": ["O", "O", "O", "I", "I", "O"], "words": ["what", "artist", "creates", "riot", "grrrl", "music"]}
input:  {'sentence': 'what country is rafael sorkin from'}
prediction:  {"logits": [[10.093888282775879, -10.376049995422363], [9.874981880187988, -10.358308792114258], [8.756714820861816, -8.979889869689941], [-10.959955215454102, 11.479487419128418], [-7.501682281494141, 7.357836723327637], [8.452631950378418, -8.934502601623535]], "class_probabilities": [[1.0, 1.2883052402656858e-09], [1.0, 1.632274093665842e-09], [1.0, 1.9819397678588757e-08], [1.7975224086974606e-10, 1.0], [3.520405869039678e-07, 0.9999996423721313], [1.0, 2.8110166638839473e-08]], "tags": ["O", "O", "O", "I", "I", "O"], "words": ["what", "country", "is", "rafael", "sorkin", "from"]}
input:  {'sentence': "what was marcy rae's profession"}
prediction:  {"logits": [[8.304543495178223, -8.49860954284668], [8.531720161437988, -8.779454231262207], [-8.729695320129395, 9.029540061950684], [-6.539294242858887, 7.029804229736328], [-5.03172492980957, 5.170454978942871], [5.006085395812988, -5.221122741699219]], "class_probabilities": [[0.9999999403953552, 5.040607931050545e-08], [0.9999999403953552, 3.0328607181218104e-08], [1.9375908877350412e-08, 1.0], [1.279425077882479e-06, 0.9999987483024597], [3.7088004319230095e-05, 0.9999629259109497], [0.9999638199806213, 3.6171306419419125e-05]], "tags": ["O", "O", "I", "I", "I", "O"], "words": ["what", "was", "marcy", "rae", "'s", "profession"]}
input:  {'sentence': 'what is the language of the invaders'}
prediction:  {"logits": [[9.052947044372559, -9.285898208618164], [9.477446556091309, -9.777214050292969], [8.2933988571167, -8.622628211975098], [5.522915840148926, -5.844305992126465], [5.893302917480469, -6.365967750549316], [-5.722286224365234, 5.7159504890441895], [-6.212862014770508, 6.36495304107666]], "class_probabilities": [[1.0, 1.0852782672543526e-08], [1.0, 4.34316893560549e-09], [0.9999999403953552, 4.502593498045826e-08], [0.9999884366989136, 1.1568398804229219e-05], [0.999995231628418, 4.740938948089024e-06], [1.0775375812954735e-05, 0.9999892115592957], [3.447647259235964e-06, 0.999996542930603]], "tags": ["O", "O", "O", "O", "O", "I", "I"], "words": ["what", "is", "the", "language", "of", "the", "invaders"]}
input:  {'sentence': 'which artist recorded srv shuffle'}
prediction:  {"logits": [[9.028275489807129, -9.2664155960083], [9.358494758605957, -9.725643157958984], [9.082541465759277, -9.613968849182129], [-7.875382900238037, 8.00829792022705], [-8.94989013671875, 9.334349632263184]], "class_probabilities":[[1.0, 1.1342704553385374e-08], [1.0, 5.150680770782401e-09], [1.0, 7.589422423848191e-09], [1.2641692137549398e-07, 0.9999998807907104], [1.1461861681993923e-08, 1.0]], "tags": ["O", "O", "O", "I", "I"], "words": ["which", "artist", "recorded", "srv", "shuffle"]}
input:  {'sentence': 'where was fran drescher born'}
prediction:  {"logits": [[8.754790306091309, -9.030854225158691], [9.944942474365234, -10.290895462036133], [-8.262197494506836, 8.387333869934082], [-9.28039264678955, 9.29737377166748], [6.310794830322266, -6.684662818908691]], "class_probabilities": [[1.0, 1.8870904838763636e-08], [1.0, 1.628123080799071e-09], [5.87759920733788e-08, 0.9999999403953552], [8.546308549739479e-09, 1.0], [0.9999977350234985, 2.2706146864948096e-06]], "tags": ["O", "O", "I", "I", "O"], "words": ["where", "was", "fran", "drescher", "born"]}
input:  {'sentence': 'what music genre is lonely revolutions'}
prediction:  {"logits": [[9.748954772949219, -9.984620094299316], [9.1851224899292, -9.520463943481445], [10.01893138885498, -10.451571464538574], [9.41489315032959, -9.791045188903809], [-8.951233863830566, 8.959811210632324], [-6.168938636779785, 6.439537048339844]], "class_probabilities": [[1.0, 2.690405453975586e-09], [1.0, 7.52085860256102e-09], [1.0, 1.2875780441845563e-09], [1.0, 4.560022137667374e-09], [1.6646845679701983e-08, 1.0], [3.3435446766816312e-06, 0.9999966621398926]], "tags":["O", "O", "O", "O", "I", "I"], "words": ["what", "music", "genre", "is", "lonely", "revolutions"]}
input:  {'sentence': 'souad massi was the music contributor for which film'}
prediction:  {"logits": [[-6.130303382873535, 6.447381496429443], [-7.978407859802246, 7.916427135467529], [7.276904106140137, -7.79086971282959], [5.6424736976623535, -6.131288528442383], [5.850767612457275, -6.345029830932617], [6.778904914855957,-7.194135665893555], [7.894515514373779, -8.433209419250488], [7.289031505584717, -7.816277027130127], [7.001927852630615, -7.390376091003418]], "class_probabilities": [[3.4480979138606926e-06, 0.999996542930603], [1.2501467949732614e-07, 0.9999998807907104], [0.9999997019767761, 2.8585699851646496e-07], [0.999992311000824, 7.704011295572855e-06], [0.9999949336051941, 5.0516177907411475e-06], [0.9999991655349731, 8.542504019715125e-07], [0.9999999403953552, 8.108850835242265e-08], [0.9999997019767761, 2.753263288468588e-07], [0.999999463558197, 5.616965381705086e-07]], "tags": ["I", "I", "O", "O", "O", "O", "O", "O", "O"], "words": ["souad", "massi", "was", "the", "music", "contributor", "for", "which", "film"]}
input:  {'sentence': 'what is the name of a title found under the thriller ( genre )'}
prediction:  {"logits": [[8.699601173400879, -9.0216703414917], [8.270527839660645, -8.536103248596191], [9.355738639831543, -9.68509578704834], [8.131817817687988, -8.393900871276855], [8.405546188354492, -8.701436996459961], [8.42676067352295, -8.60349178314209], [8.879528045654297, -9.043085098266602], [9.207085609436035, -9.324575424194336], [9.0308198928833, -9.31006145477295], [7.66093635559082, -8.103529930114746], [-8.905280113220215, 8.73732852935791], [5.728875637054443, -6.01603364944458], [6.5330657958984375, -7.094614028930664], [4.630368709564209, -5.1072306632995605]], "class_probabilities": [[1.0, 2.0125634492274003e-08], [0.9999999403953552, 5.023111526725188e-08], [1.0, 5.378617551343723e-09], [0.9999999403953552, 6.65229507035292e-08], [0.9999999403953552, 3.719902963439381e-08], [0.9999999403953552, 4.0165698322880417e-08], [1.0, 1.6455382834124066e-08], [1.0, 8.94955576313805e-09], [1.0, 1.0830698116137683e-08], [0.9999998807907104, 1.42422706517209e-07], [2.1772706304545864e-08, 1.0], [0.9999920725822449, 7.929526873340365e-06], [0.9999988079071045, 1.2066277577105211e-06], [0.9999409914016724, 5.901857002754696e-05]], "tags": ["O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "I", "O", "O", "O"], "words": ["what", "is", "the", "name", "of", "a", "title", "found", "under", "the", "thriller", "(", "genre", ")"]}
input:  {'sentence': 'what is a book from the ethan frome series'}
prediction:  {"logits": [[8.339886665344238, -8.62112045288086], [8.4562406539917, -8.724736213684082], [8.209368705749512, -8.512554168701172], [7.625807285308838, -7.918526649475098], [6.5768351554870605, -6.826633453369141], [4.668245792388916, -5.002597808837891], [-4.254884719848633, 4.286087989807129], [-5.124143600463867, 4.996259689331055], [5.994084358215332, -6.146832466125488]], "class_probabilities": [[0.9999999403953552, 4.3045581321621285e-08], [0.9999999403953552, 3.454590213891606e-08], [0.9999999403953552, 5.467146735327333e-08], [0.9999998211860657, 1.7749302116953913e-07], [0.9999985098838806, 1.5098947869773838e-06], [0.9999368786811829, 6.309265154413879e-05], [0.0001952620514202863, 0.9998047351837158], [4.0248269215226173e-05, 0.9999597668647766], [0.9999946355819702, 5.336597951099975e-06]], "tags": ["O", "O", "O", "O", "O", "O", "I", "I", "O"], "words": ["what", "is", "a", "book", "from", "the", "ethan", "frome", "series"]}
input:  {'sentence': 'where is look from'}
prediction:  {"logits": [[8.6942720413208, -8.996464729309082], [5.268302917480469, -5.263429164886475], [-9.595760345458984, 9.643467903137207], [6.649141788482666, -6.9167633056640625]], "class_probabilities": [[1.0, 2.074964378095956e-08], [0.9999732971191406, 2.6675654225982726e-05], [4.410722453940252e-09, 1.0], [0.9999986886978149, 1.2835179177272948e-06]], "tags": ["O", "O", "I", "O"], "words": ["where", "is", "look", "from"]}
input:  {'sentence': 'who is the producer on evangelion : 3.0 you can ( not ) redo'}
prediction:  {"logits": [[7.11078405380249, -7.239439010620117], [7.310622215270996, -7.488116264343262], [8.282651901245117, -8.624699592590332], [6.319696426391602, -6.70600700378418], [7.473918914794922, -7.858055591583252], [-8.751923561096191, 8.641980171203613], [-8.525031089782715, 8.855090141296387], [-8.940231323242188, 9.346726417541504], [-9.02168083190918, 9.426438331604004], [-8.78761100769043, 9.333992958068848], [-6.7477521896362305, 7.227974891662598], [-7.216712474822998, 7.74497127532959], [-6.859787464141846, 7.457984924316406], [-5.099355220794678, 5.641816139221191]], "class_probabilities": [[0.9999994039535522, 5.858370855094108e-07], [0.9999996423721313, 3.741014324987191e-07], [0.9999999403953552, 4.5418218519444054e-08], [0.9999977946281433, 2.2029666979506146e-06], [0.9999997615814209, 2.194866937088591e-07], [2.792052633537878e-08, 1.0], [2.8308004829113997e-08, 1.0], [1.143077277276916e-08, 1.0], [9.729336447605874e-09, 1.0], [1.3486118888295096e-08, 1.0], [8.519585890098824e-07, 0.9999991655349731], [3.1785083365321043e-07, 0.9999997019767761], [6.051602667866973e-07, 0.9999994039535522], [2.163512363040354e-05, 0.9999783635139465]], "tags": ["O", "O", "O", "O", "O", "I", "I", "I", "I", "I", "I", "I","I", "I"], "words": ["who", "is", "the", "producer", "on", "evangelion", ":", "3.0", "you", "can", "(", "not", ")", "redo"]}
input:  {'sentence': "what medium was smooth ' n swingin ' released on"}
prediction:  {"logits": [[7.0971221923828125, -7.196249961853027], [6.718428611755371, -7.002164840698242], [8.246155738830566, -8.642882347106934], [-5.272072792053223, 5.2099409103393555], [-4.397397994995117, 4.482825756072998], [-8.607684135437012, 8.945131301879883], [-9.585654258728027, 10.044259071350098], [-7.419056415557861, 7.750012397766113], [1.526052474975586, -1.757972240447998], [3.918844223022461, -4.212307929992676]], "class_probabilities": [[0.9999994039535522, 6.201076416800788e-07], [0.999998927116394, 1.0995664752044831e-06], [0.9999999403953552, 4.6257689234607824e-08], [2.80354233836988e-05, 0.9999719858169556], [0.00013909360859543085, 0.9998608827590942], [2.3818232719463595e-08, 1.0], [2.9842621707132366e-09, 1.0], [2.583192895144748e-07, 0.9999997615814209], [0.9638766646385193, 0.03612332046031952], [0.9997058510780334, 0.00029414246091619134]], "tags": ["O", "O", "O", "I", "I", "I", "I", "I", "O", "O"], "words": ["what", "medium", "was", "smooth", "'", "n", "swingin", "'", "released", "on"]}
input:  {'sentence': "what was bill hosket , jr . ' s position"}
prediction:  {"logits": [[7.605815887451172, -7.657910346984863], [8.589877128601074, -8.853836059570312], [-9.757682800292969, 9.936151504516602], [-9.453827857971191, 9.887608528137207], [-6.876792907714844, 7.352346897125244], [-6.683572292327881, 6.554892539978027], [7.496373176574707, -8.074542045593262], [7.126738548278809, -7.606776237487793], [6.740769386291504, -7.229009628295898], [7.759803771972656, -8.210601806640625]], "class_probabilities": [[0.9999997615814209, 2.349891587982711e-07], [1.0, 2.656386044463943e-08], [2.7994739859593665e-09, 1.0], [3.9821803632378305e-09, 1.0], [6.612459060306719e-07, 0.9999993443489075], [1.7807674339564983e-06, 0.9999982118606567], [0.9999998211860657, 1.7283726094774465e-07], [0.9999995827674866, 3.993150130554568e-07], [0.9999991655349731, 8.570411864639027e-07], [0.9999998807907104, 1.1591534132548986e-07]], "tags": ["O", "O", "I", "I", "I", "I", "O", "O", "O", "O"], "words": ["what", "was", "bill", "hosket", ",", "jr", ".", "'", "s","position"]}
input:  {'sentence': 'what did oscar saul die from'}
prediction:  {"logits": [[9.195761680603027, -9.47659683227539], [8.394570350646973, -8.637711524963379], [-9.209511756896973, 9.387480735778809], [-8.177053451538086, 8.113224029541016], [5.387872695922852, -5.906167984008789], [8.295058250427246, -8.761612892150879]], "class_probabilities": [[1.0, 7.774945132155153e-09], [0.9999999403953552, 4.008427012536231e-08], [8.383565841540985e-09, 1.0], [8.41826022224268e-08, 0.9999999403953552], [0.9999875426292419, 1.2446724213077687e-05], [0.9999999403953552, 3.911846846449407e-08]], "tags": ["O", "O", "I", "I", "O", "O"], "words": ["what", "did", "oscar", "saul", "die", "from"]}
input:  {'sentence': 'what industry does critical software operate in'}
prediction:  {"logits": [[9.233717918395996, -9.46999454498291], [9.049087524414062, -9.398883819580078], [9.560789108276367, -9.976393699645996], [-7.223443508148193, 7.322834014892578], [-6.649218559265137, 6.440572261810303], [7.066544532775879, -7.683723449707031], [7.797279357910156, -8.271686553955078]], "class_probabilities": [[1.0, 7.53495843497376e-09], [1.0, 9.730784178429985e-09], [1.0, 3.2742342170877237e-09], [4.815390752810345e-07, 0.9999995231628418], [2.066214392471011e-06, 0.9999979138374329], [0.9999995827674866, 3.9268095974875905e-07], [0.9999998807907104, 1.05035645958651e-07]], "tags": ["O", "O", "O", "I", "I", "O", "O"], "words": ["what", "industry", "does", "critical", "software", "operate", "in"]}
input:  {'sentence': 'what is the active ingredient in titanium dioxide 0.223 cream'}
prediction:  {"logits": [[8.705452919006348, -8.929789543151855], [8.374844551086426, -8.701600074768066], [9.176691055297852, -9.576544761657715], [8.361573219299316, -8.766791343688965], [8.242751121520996, -8.692540168762207], [8.850594520568848,-9.32579231262207], [-8.92741870880127, 8.9033842086792], [-9.251662254333496, 9.544903755187988], [-8.596491813659668, 9.045969009399414], [-6.699119567871094, 7.281694412231445]], "class_probabilities": [[1.0, 2.1933679761332314e-08], [0.9999999403953552, 3.8352556686049866e-08], [1.0, 7.170898541630777e-09], [0.9999999403953552, 3.64121035545395e-08], [0.9999999403953552, 4.416684973307383e-08], [1.0, 1.2767182866468829e-08], [1.8037679794247197e-08, 1.0], [6.866811119721206e-09, 1.0], [2.177594637942093e-08, 1.0], [8.476357606923557e-07, 0.9999991655349731]], "tags": ["O", "O", "O", "O", "O", "O", "I", "I", "I", "I"], "words": ["what", "is", "the", "active", "ingredient", "in", "titanium", "dioxide", "0.223", "cream"]}
input:  {'sentence': 'what is the active ingredient moiety in fast freeze'}
prediction:  {"logits": [[8.89854907989502, -9.202943801879883], [8.707925796508789, -9.043037414550781], [9.752339363098145, -10.110350608825684], [9.884659767150879, -10.297396659851074], [10.43460750579834, -10.87729549407959], [8.152100563049316, -8.680243492126465], [5.036672592163086, -5.245060920715332], [-9.730332374572754, 9.764254570007324], [-6.804563522338867, 6.942901611328125]], "class_probabilities": [[1.0, 1.3760084627278957e-08], [1.0, 1.9536853912427432e-08], [1.0, 2.364521911601969e-09], [1.0, 1.7180834532837252e-09], [1.0, 5.550832571366016e-10], [0.9999999403953552, 4.895598948451152e-08], [0.9999657273292542, 3.425192699069157e-05], [3.416712690551549e-09, 1.0], [1.0704128499128274e-06, 0.999998927116394]], "tags": ["O", "O", "O", "O", "O", "O", "O", "I", "I"], "words": ["what", "is", "the", "active", "ingredient", "moiety", "in", "fast", "freeze"]}
input:  {'sentence': 'what did sheldon breiner invent'}
prediction:  {"logits": [[8.891865730285645, -9.099478721618652], [7.4920854568481445, -7.699904441833496], [-8.97818374633789, 9.231513023376465], [-7.9236555099487305, 8.138079643249512], [5.372681617736816, -5.583972930908203]], "class_probabilities": [[1.0, 1.5362376259986377e-08], [0.9999997615814209, 2.5246578161386424e-07], [1.2348913891457869e-08, 1.0], [1.0579788778386501e-07, 0.9999998807907104], [0.9999825358390808, 1.7441258023609407e-05]], "tags": ["O", "O", "I", "I", "O"], "words": ["what", "did", "sheldon", "breiner", "invent"]}
input:  {'sentence': 'what entity was involved in the roman – parthian war of 58 – 63'}
prediction:  {"logits": [[9.015543937683105, -9.283906936645508], [7.80557918548584, -8.151385307312012], [10.009781837463379, -10.288052558898926], [4.923876762390137, -5.183717727661133], [6.049535274505615, -6.348017692565918], [5.702660083770752, -6.099062919616699], [-7.098887920379639, 7.061542510986328], [-8.12786865234375, 8.339319229125977], [-7.342863082885742, 7.733397006988525], [-7.558258056640625, 8.048397064208984], [-7.500312805175781, 8.090081214904785], [-7.6976318359375, 8.327300071716309], [-8.42745304107666, 9.016975402832031], [-6.072296142578125, 6.498757362365723]], "class_probabilities": [[1.0, 1.1288854295798956e-08], [0.9999998807907104, 1.1748389283638971e-07], [1.0, 1.5302504818848206e-09], [0.9999592304229736,4.076709592482075e-05], [0.9999958872795105, 4.1286643863713834e-06], [0.9999924898147583, 7.491579253837699e-06], [7.08276274963282e-07, 0.9999992847442627], [7.053279915680832e-08, 0.9999999403953552], [2.8344155111881264e-07, 0.9999997019767761],[1.6676915493007982e-07, 0.9999998211860657], [1.6950316705788282e-07, 0.9999998211860657], [1.0976402364804017e-07, 0.9999998807907104], [2.6544919151660906e-08, 1.0], [3.4710376439761603e-06, 0.999996542930603]], "tags": ["O", "O", "O", "O", "O", "O", "I", "I", "I", "I", "I", "I", "I", "I"], "words": ["what", "entity", "was", "involved", "in", "the", "roman", "\u2013", "parthian", "war", "of", "58", "\u2013", "63"]}
input:  {'sentence': 'what is the name for meccas canonical version'}
prediction:  {"logits": [[4.055498123168945, -4.273970603942871], [4.68829345703125, -4.788795471191406], [5.581427574157715, -5.8751325607299805], [4.417109489440918, -4.698109149932861], [5.517838001251221, -5.70955228805542], [-4.859011650085449,4.842243671417236], [6.559183597564697, -6.679147720336914], [5.939227104187012, -6.231776714324951]], "class_probabilities": [[0.999758780002594, 0.00024124199990183115], [0.9999234080314636, 7.658069807803258e-05], [0.9999894499778748, 1.057972895068815e-05], [0.9998900294303894, 0.0001099671280826442], [0.9999867081642151, 1.3304561434779316e-05], [6.120283796917647e-05, 0.9999387860298157], [0.9999982118606567, 1.7810051531341742e-06], [0.9999948143959045, 5.17843045599875e-06]], "tags": ["O", "O", "O", "O", "O", "I", "O", "O"], "words": ["what", "is", "the", "name", "for", "meccas", "canonical", "version"]}
input:  {'sentence': 'what country in north america released the film third man out'}
prediction:  {"logits": [[8.383902549743652, -8.653858184814453], [8.405646324157715, -8.77161979675293], [9.487135887145996, -9.95664119720459], [4.605401515960693, -4.886800765991211], [6.048155784606934, -6.270119667053223], [8.510970115661621, -8.7835693359375], [8.026629447937012, -8.489107131958008], [7.98345422744751, -8.461077690124512], [-8.365424156188965, 8.300347328186035], [-7.838351249694824, 8.037996292114258], [-6.414721488952637, 6.723383903503418]], "class_probabilities": [[0.9999999403953552, 3.986521690535483e-08], [0.9999999403953552, 3.467436471282781e-08], [1.0, 3.594801345485621e-09], [0.9999245405197144, 7.543205720139667e-05], [0.9999955296516418, 4.469294708542293e-06], [0.9999999403953552, 3.083731314745819e-08], [0.9999999403953552, 6.719038481151074e-08], [0.9999999403953552, 7.214899255814089e-08], [5.782922585240158e-08, 0.9999999403953552], [1.2734732024455298e-07, 0.9999998807907104], [1.96875862457091e-06, 0.9999980330467224]], "tags": ["O", "O", "O", "O", "O", "O", "O", "O", "I", "I", "I"], "words": ["what", "country", "in", "north", "america", "released", "the", "film", "third", "man", "out"]}
input:  {'sentence': 'what system is used for electric power distribution'}
prediction:  {"logits": [[9.256540298461914, -9.561306953430176], [6.136517524719238, -6.518344879150391], [10.73792839050293, -10.982171058654785], [9.495895385742188, -9.883383750915527], [7.6006574630737305, -8.025996208190918], [-7.833144664764404, 7.782502174377441], [-6.796451091766357, 7.146385192871094], [-5.916768550872803, 6.31689453125]], "class_probabilities": [[1.0, 6.722214340726396e-09], [0.9999967813491821, 3.191991254425375e-06], [1.0, 3.6904579392427195e-10], [1.0, 3.834295547733291e-09], [0.9999998211860657, 1.634671349393102e-07], [1.6527641832908557e-07, 0.9999998211860657], [8.804456115285575e-07, 0.9999991059303284], [4.863907633989584e-06, 0.9999951124191284]], "tags": ["O", "O", "O", "O", "O", "I", "I", "I"], "words": ["what", "system", "is", "used", "for", "electric", "power", "distribution"]}
input:  {'sentence': 'what is a live action action comedy film produce by spyglass entertainment'}
prediction:  {"logits": [[7.752961158752441, -7.833697319030762], [7.772695541381836, -7.8983540534973145], [7.923630714416504, -8.111820220947266], [6.936495780944824, -7.117203235626221], [6.584548473358154, -6.7446184158325195], [6.176007270812988, -6.237911224365234], [6.4084062576293945, -6.383807182312012], [6.864221572875977, -6.716526985168457], [6.951045513153076, -6.967777252197266], [4.743659019470215, -4.499826908111572], [-7.8611273765563965, 8.29308032989502], [-5.709151268005371,5.837172031402588]], "class_probabilities": [[0.9999998211860657, 1.7013753961236944e-07], [0.9999998211860657, 1.5636867090051965e-07], [0.9999998807907104, 1.0861567290021412e-07], [0.9999992251396179, 7.880531711634831e-07], [0.9999983906745911, 1.6263569477814599e-06], [0.9999959468841553, 4.061645086039789e-06], [0.9999971985816956, 2.78234574579983e-06], [0.9999987483024597, 1.264606112272304e-06], [0.9999991059303284, 9.018441460284521e-07], [0.9999032616615295, 9.673037129687145e-05], [9.645325604878963e-08, 0.9999998807907104], [9.671438419900369e-06, 0.9999903440475464]], "tags": ["O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "I", "I"], "words": ["what", "is", "a", "live", "action", "action", "comedy", "film", "produce", "by", "spyglass", "entertainment"]}
input:  {'sentence': 'which genre is teri meri kahaani under'}
prediction:  {"logits": [[8.592487335205078, -8.861180305480957], [8.438044548034668, -8.851397514343262], [9.69798469543457, -10.11215877532959], [-7.054055690765381, 7.154474258422852], [-9.467647552490234, 10.020216941833496], [-7.969315528869629, 7.976597785949707], [7.089127540588379, -7.613999366760254]], "class_probabilities": [[1.0, 2.630079443122213e-08], [0.9999999403953552, 3.099493284253185e-08], [1.0, 2.492094086647967e-09], [6.7501486000765e-07, 0.9999993443489075], [3.439755591472249e-09, 1.0], [1.1878942984822061e-07, 0.9999998807907104], [0.9999995827674866, 4.116355967198615e-07]], "tags": ["O", "O", "O", "I", "I", "I", "O"], "words": ["which", "genre", "is", "teri", "meri", "kahaani", "under"]}
input:  {'sentence': 'what is the genre of the artists weird war'}
prediction:  {"logits": [[8.874188423156738, -9.318450927734375], [8.611854553222656, -8.932538986206055], [9.833763122558594, -10.256089210510254], [9.603140830993652, -9.983777046203613], [9.438742637634277, -9.839436531066895], [8.56116008758545,-9.02134895324707], [5.292727470397949, -5.6274333000183105], [-8.025317192077637, 8.095510482788086], [-7.340813636779785, 7.747504711151123]], "class_probabilities": [[1.0, 1.2561384821196953e-08], [1.0, 2.401965204512635e-08], [1.0, 1.884032485577336e-09], [1.0, 3.1153701840480608e-09], [1.0, 4.242219908690004e-09], [1.0, 2.3121380365864752e-08], [0.9999819397926331, 1.808950946724508e-05], [9.972725933948823e-08, 0.9999998807907104], [2.800442473471776e-07, 0.9999997019767761]], "tags": ["O", "O", "O", "O", "O", "O", "O", "I", "I"], "words": ["what", "is", "the", "genre", "of", "the", "artists", "weird", "war"]}
input:  {'sentence': "who wrote gulliver's travels"}
prediction:  {"logits": [[7.784527778625488, -8.078686714172363], [9.186930656433105, -9.534187316894531], [-9.436686515808105, 9.504347801208496], [-8.898358345031738, 9.213709831237793], [-6.991088390350342, 7.3979692459106445]], "class_probabilities": [[0.9999998807907104, 1.2903080914838938e-07], [1.0, 7.404936663846229e-09], [5.943103786876236e-09, 1.0], [1.3615347960183044e-08, 1.0], [5.635229740619252e-07, 0.999999463558197]], "tags": ["O", "O", "I", "I", "I"], "words": ["who", "wrote", "gulliver", "'s", "travels"]}
input:  {'sentence': "what's the name of a founder of ndtv"}
prediction:  {"logits": [[9.722207069396973, -10.016923904418945], [8.228691101074219, -8.532501220703125], [7.916787147521973, -8.331719398498535], [8.07663631439209, -8.405237197875977], [9.168102264404297, -9.591585159301758], [8.399558067321777,-8.726409912109375], [7.659846782684326, -7.900608062744141], [7.565478801727295, -7.873752593994141], [-7.281405448913574, 7.5517449378967285]], "class_probabilities": [[1.0, 2.675493604442636e-09], [0.9999999403953552, 5.2566207386917085e-08], [0.9999999403953552, 8.777346494071026e-08], [0.9999999403953552, 6.950461539645403e-08], [1.0, 7.124776768563379e-09], [0.9999999403953552, 3.649943991490545e-08], [0.9999998211860657, 1.7465460189214355e-07], [0.9999998211860657, 1.9716358679033874e-07], [3.614466947965411e-07, 0.9999996423721313]], "tags": ["O", "O", "O", "O", "O", "O", "O", "O", "I"], "words": ["what", "'s", "the", "name", "of", "a", "founder", "of", "ndtv"]}
Traceback (most recent call last):
  File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/usr/lib/python3.6/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/run.py", line 13, in <module>
    main(prog="python -m allennlp.run")
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/__init__.py", line 77, in main
    args.func(args)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/predict.py", line 145, in predict_inner
    _run(predictor, input_file, output_file, args.batch_size, not args.silent, args.cuda_device)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/predict.py", line 120, in _run
    _run_predictor(batch_json_data)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/predict.py", line 98, in _run_predictor
    result = predictor.predict_json(batch_data[0], cuda_device)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/service/predictors/sentence_tagger.py", line 40, in predict_json
    output = self._model.forward_on_instance(instance, cuda_device)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/models/model.py", line 117, in forward_on_instance
    outputs = self.decode(self.forward(**model_input))
  File "/home/michael/Desktop/lattice/allennlp/allennlp/models/simple_tagger.py", line 99, in forward
    embedded_text_input = self.text_field_embedder(tokens)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/text_field_embedders/basic_text_field_embedder.py", line 47, in forward
    token_vectors = embedder(tensor)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/token_embedders/token_characters_encoder.py", line 36, in forward
    return self._dropout(self._encoder(self._embedding(token_characters), mask))
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/time_distributed.py", line 35, in forward
    reshaped_outputs = self._module(*reshaped_inputs)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/seq2vec_encoders/cnn_encoder.py", line 103, in forward
    for convolution_layer in self._convolution_layers]
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/seq2vec_encoders/cnn_encoder.py", line 103, in <listcomp>
    for convolution_layer in self._convolution_layers]
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/conv.py", line 154, in forward
    self.padding, self.dilation, self.groups)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/functional.py", line 83, in conv1d
    return f(input, weight, bias)
RuntimeError: Given input size: (8 x 1 x 4). Calculated output size: (50 x 1 x 0). Output size is too small at /pytorch/torch/lib/THNN/generic/SpatialConvolutionMM.c:45

[matt-gardner]

Yes, this is a known issue, unfortunately. The problem is that you're specifying 5-grams for your CNN filters, but you have no word that is over 4 characters, so the CNN filter crashes. Detecting this and fixing it on our side is actually quite hard, but it's relatively easy to fix in your configuration. Either you can change the size of the filters in your CNN, or add padding around the characters so that you'll always have at least five characters (I'd recommend the second option). You can see how we do that with BiDAF here: https://github.com/allenai/allennlp/blob/83f0c5ecaa1020371d7788f6683764ff806dbe36/training_config/bidaf.json#L9-L16

You could just do "end_tokens": [0, 0, 0, 0], making sure there's always enough 0 padding, or you could put explicit begin and end markers, like the BiDAF model does.

[PetrochukM]

Added this line:

              "character_tokenizer": { 
                "byte_encoding": "utf-8",
                "end_tokens": [0, 0, 0, 0] 
              },

I am assuming that I need to retrain the model. The error still occurs.

Using tutorials/getting_started/simple_tagger.json.

[matt-gardner]

You shouldn't have to retrain the model (unless you weren't using byte encoding before - you can do this without byte encoding by changing [0, 0, 0, 0] to ["@@PADDING@@", ...]), but you do need to update the configuration that's inside the model archive.

[PetrochukM]

Hi!

I tried to add those lines but then I get this weird error if I try to retrain with byte encoding. Without byte encoding and "@@PADDING@@" everything works!

  0%|          | 0/330 [00:00<?, ?it/s]2017-12-04 14:17:01,985 - INFO - allennlp.training.trainer - Training
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [32,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [33,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [34,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [35,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [36,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [37,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [38,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [39,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [40,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [41,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [42,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [43,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [44,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [45,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [46,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
/pytorch/torch/lib/THC/THCTensorIndex.cu:321: void indexSelectLargeIndex(TensorInfo<T, IndexType>, TensorInfo<T, IndexType>, TensorInfo<long, IndexType>, int, int, IndexType, IndexType, long) [with T = float, IndexType = unsigned int, DstDim = 2, SrcDim = 2, IdxDim = -2]: block: [33,0,0], thread: [47,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
Traceback (most recent call last):
  File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/usr/lib/python3.6/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/run.py", line 13, in <module>
    main(prog="python -m allennlp.run")
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/__init__.py", line 77, in main
    args.func(args)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/train.py", line 73, in train_model_from_args
    train_model_from_file(args.param_path, args.serialization_dir)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/train.py", line 89, in train_model_from_file
    return train_model(params, serialization_dir)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/commands/train.py", line 178, in train_model
    trainer.train()
  File "/home/michael/Desktop/lattice/allennlp/allennlp/training/trainer.py", line 369, in train
    train_metrics = self._train_epoch(epoch)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/training/trainer.py", line 221, in _train_epoch
    loss = self._batch_loss(batch, for_training=True)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/training/trainer.py", line 176, in _batch_loss
    output_dict = self._forward(batch, for_training=for_training)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/training/trainer.py", line 410, in _forward
    return self._model.forward(**tensor_batch)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/models/simple_tagger.py", line 99, in forward
    embedded_text_input = self.text_field_embedder(tokens)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/text_field_embedders/basic_text_field_embedder.py", line 47, in forward
    token_vectors = embedder(tensor)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/token_embedders/token_characters_encoder.py", line 36, in forward
    return self._dropout(self._encoder(self._embedding(token_characters), mask))
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/time_distributed.py", line 35, in forward
    reshaped_outputs = self._module(*reshaped_inputs)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/seq2vec_encoders/cnn_encoder.py", line 103, in forward
    for convolution_layer in self._convolution_layers]
  File "/home/michael/Desktop/lattice/allennlp/allennlp/modules/seq2vec_encoders/cnn_encoder.py", line 103, in <listcomp>
    for convolution_layer in self._convolution_layers]
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 224, in __call__
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/conv.py", line 154, in forward
    self.padding, self.dilation, self.groups)
  File "/usr/local/lib/python3.6/dist-packages/torch/nn/functional.py", line 83, in conv1d
    return f(input, weight, bias)
RuntimeError: CUDNN_STATUS_NOT_INITIALIZED

[matt-gardner]

If you want to use byte encoding, you also need to specify the number of embeddings in your character embedder: https://github.com/allenai/allennlp/blob/3995f70c1cb6190352eb5e063e3ba37b3121112f/training_config/bidaf.json#L31-L35

This is because you're no longer using the vocabulary to determine how many characters there are - the code will think you want an embedding matrix with zero entries, and things will fail, as you see.

[PetrochukM]

Okay! Thanks for your help!

[matt-gardner]

No problem!

[matt-gardner]

Oh, for anyone stumbling upon this issue later, here's another workaround that I should have suggested first:

python -m allennlp.run predict /tmp/subject_recognition/model.tar.gz INPUT_FILE --overrides="dataset_reader.token_indexers.token_characters.character_tokenizer.end_tokens = ['@@PADDING@@', '@@PADDING@@']"

That is, the problem was with the fact that there wasn't enough padding on the characters. There's an option in the predict command to override components of the configuration file. One of the items in the configuration file lets you add padding when tokenizing the characters. So you can just run the original command with the override, and it should fix the problem.