-
Notifications
You must be signed in to change notification settings - Fork 0
/
predict.py
356 lines (274 loc) · 15.4 KB
/
predict.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
import re
import os
import json
import os
import copy
import pandas as pd
import numpy as np
import tensorflow.keras as K
from collections import namedtuple
from tqdm import tqdm, tnrange
from nltk.corpus import wordnet as wn
import utils
import parsers
import models
import generatorPrototype
import generatorMultitask
def predict_babelnet(input_path : str, output_path : str, resources_path : str) -> None:
"""
DO NOT MODIFY THE SIGNATURE!
This is the skeleton of the prediction function.
The predict function will build your model, load the weights from the checkpoint and write a new file (output_path)
with your predictions in the "<id> <BABELSynset>" format (e.g. "d000.s000.t000 bn:01234567n").
The resources folder should contain everything you need to make the predictions. It is the "resources" folder in your submission.
N.B. DO NOT HARD CODE PATHS IN HERE. Use resource_path instead, otherwise we will not be able to run the code.
If you don't know what HARD CODING means see: https://en.wikipedia.org/wiki/Hard_coding
:param input_path: the path of the input file to predict in the same format as Raganato's framework (XML files you downloaded).
:param output_path: the path of the output file (where you save your predictions)
:param resources_path: the path of the resources folder containing your model and stuff you might need.
:return: None
"""
main_predict_multitask(input_path, output_path, resources_path, prediction_type='babelnet')
def predict_wordnet_domains(input_path : str, output_path : str, resources_path : str) -> None:
"""
DO NOT MODIFY THE SIGNATURE!
This is the skeleton of the prediction function.
The predict function will build your model, load the weights from the checkpoint and write a new file (output_path)
with your predictions in the "<id> <wordnetDomain>" format (e.g. "d000.s000.t000 sport").
The resources folder should contain everything you need to make the predictions. It is the "resources" folder in your submission.
N.B. DO NOT HARD CODE PATHS IN HERE. Use resource_path instead, otherwise we will not be able to run the code.
If you don't know what HARD CODING means see: https://en.wikipedia.org/wiki/Hard_coding
:param input_path: the path of the input file to predict in the same format as Raganato's framework (XML files you downloaded).
:param output_path: the path of the output file (where you save your predictions)
:param resources_path: the path of the resources folder containing your model and stuff you might need.
:return: None
"""
main_predict_multitask(input_path, output_path, resources_path, prediction_type='wordnet_domains')
def predict_lexicographer(input_path : str, output_path : str, resources_path : str) -> None:
"""
DO NOT MODIFY THE SIGNATURE!
This is the skeleton of the prediction function.
The predict function will build your model, load the weights from the checkpoint and write a new file (output_path)
with your predictions in the "<id> <lexicographerId>" format (e.g. "d000.s000.t000 noun.animal").
The resources folder should contain everything you need to make the predictions. It is the "resources" folder in your submission.
N.B. DO NOT HARD CODE PATHS IN HERE. Use resource_path instead, otherwise we will not be able to run the code.
If you don't know what HARD CODING means see: https://en.wikipedia.org/wiki/Hard_coding
:param input_path: the path of the input file to predict in the same format as Raganato's framework (XML files you downloaded).
:param output_path: the path of the output file (where you save your predictions)
:param resources_path: the path of the resources folder containing your model and stuff you might need.
:return: None
"""
main_predict_multitask(input_path, output_path, resources_path, prediction_type='lexicographer')
def eval_parser(path, batch_size = 64):
data_flow = parsers.TrainingParser(path)
sentence_batch = []
for batch_count, sentence in enumerate(data_flow.parse(), start = 1):
sentence_batch.append(sentence)
if len(sentence_batch)==(batch_size):#==0:
yield sentence_batch
sentence_batch = []
if len(sentence_batch)>0:
yield sentence_batch
def basic_predict(batch_ground_truth_sentences,
batch_model_predictions,
candidate_synsets,
PADDING_SIZE,
reverse_output_vocab):
"""
Peforms predictions on a batch for the basic model
:param batch_ground_truth_sentences:
:param batch_model_predictions:
:param candidate_synsets candidate_synsets:
:param PADDING_SIZE:
:param reverse_output_vocab:
return predicted: batch of tuple(Sentence_id, WordNet)
"""
outputs = []
output = namedtuple("output", "Sentence_id WordNet success")
for idx_sentence, sentence in enumerate(batch_model_predictions):
ground_truth_sentence = batch_ground_truth_sentences[idx_sentence]
for idx, entry in enumerate(ground_truth_sentence):
if entry.instance == True: #only for instances not wf
if idx<PADDING_SIZE:
#WSD argmax
word_prob = sentence[idx]
current_candidate_synsets = candidate_synsets[idx_sentence][idx]
prob_dist_candidate_synset = word_prob[current_candidate_synsets]
current_synset = np.argmax(prob_dist_candidate_synset)
if current_synset>4: #change after deleting start stop
item = output(Sentence_id = entry.id_, WordNet = reverse_output_vocab[current_synset], success = True)
outputs.append(item)
else: #fallback
word = entry.lemma
item = output(Sentence_id = entry.id_, WordNet = models.MFS.retrieve_item(word), success = False)
outputs.append(item)
else: #predict truncated
word = entry.lemma
item = output(Sentence_id = entry.id_, WordNet = models.MFS.retrieve_item(word), success = False)
outputs.append(item)
return outputs
def main_predict(input_path, output_path, resources_path, prediction_type, batch_size=64, PADDING_SIZE=30):
"""
:param input_path:
:param output_path:
:param resources_path:
:param prediction_type: either babelnet, wordnet_domains, or lexicographer
:param batch_size: depends on the model
:param PADDING_SIZE: depends on the model
:return: None
"""
K.backend.clear_session()
# ##################
# # vocab loading #
# #################
mapping = pd.read_csv(os.path.join(resources_path, "mapping.csv"))
senses = utils.json_vocab_reader(os.path.join(resources_path, 'semcor.vocab.WordNet.json'))
inputs, antivocab = utils.json_vocab_reader(os.path.join(resources_path, 'semcor.input.vocab.json'),
os.path.join(resources_path, 'semcor.leftout.vocab.json'))
output_vocab = utils.merge_vocabulary(senses, inputs)
reverse_output_vocab = dict((v, k) for k, v in output_vocab.items())
# ##################
# # Model loading #
# #################
model_path, model_weight_path = sorted([os.path.join(resources_path,
os.path.join('models/best_model', i)) for i in os.listdir(
os.path.join(resources_path, 'models/best_model')) if i.startswith("model")])
model_path = os.path.join(resources_path, 'models/model_2019-09-09_21:26:42_+0000.h5')
model_weight_path = os.path.join(resources_path, 'models/model_weights_2019-09-09_21:26:42_+0000.h5')
loaded_model = K.models.load_model(model_path)
loaded_model.load_weights(model_weight_path)
# ####################
# # predicting phase #
# ###################
eval_generator = generatorPrototype.get(batch_size = batch_size,
training_file_path = input_path,
antivocab = antivocab,
output_vocab = output_vocab,
PADDING_SIZE = PADDING_SIZE)
real_words = eval_parser(path = input_path, batch_size = batch_size)
#print("starting to write to:\t{}".format(output_path))
with open(output_path, mode="a") as out:
for batch_ground_truth_sentences in tqdm(real_words, desc='batch: '):
batch_x, candidate_synsets = next(eval_generator)
batch_model_predictions = loaded_model.predict_on_batch(batch_x)
batch_outputs = basic_predict(batch_ground_truth_sentences,
batch_model_predictions,
candidate_synsets,
PADDING_SIZE,
reverse_output_vocab)
for line in batch_outputs:
#retreive appropriate mapping
pred = mapping[mapping.WordNet==line.WordNet][prediction_type].values
assert len(pred)==1, "error in mapping {}" .format(line.WordNet)
fmt = "{} {} \n".format(line.Sentence_id, pred[0])
out.write(fmt)
print("done writing to:\t{}".format(output_path))
def MFS_predict_writer(input_path, output_path, resources_path, prediction_type, batch_size=64):
"""
:param input_path:
:param output_path:
:param resources_path:
:param batch_size: depends on the model
:param prediction_type: either babelnet, wordnet_domains, or lexicographer
:return: None
"""
mapping = pd.read_csv(os.path.join(resources_path, "mapping.csv"))
real_words = eval_parser(path = input_path, batch_size = batch_size)
def MFS_predict(batch_ground_truth_sentences):
"""
Peforms MFS predictions on a batch for
:param batch_ground_truth_sentences:
return predicted: batch of tuple(Sentence_id, WordNet)
"""
outputs = []
output = namedtuple("output", "Sentence_id WordNet")
for idx_sentence, sentence in enumerate(batch_ground_truth_sentences):
for idx, entry in enumerate(sentence):
if entry.instance == True: #only for instances not wf
word = entry.lemma
item = output(Sentence_id = entry.id_, WordNet = models.MFS.retrieve_item(word))
outputs.append(item)
return outputs
with open(output_path, mode="a") as out:
for batch_ground_truth_sentences in tqdm(real_words, desc='batch: '):
batch_outputs = MFS_predict(batch_ground_truth_sentences)
for line in batch_outputs:
#retreive appropriate mapping
pred = mapping[mapping.WordNet==line.WordNet][prediction_type].values
assert len(pred)==1, "error in mapping {}" .format(line.WordNet)
fmt = "{} {} \n".format(line.Sentence_id, pred[0])
out.write(fmt)
print("done writing to:\t{}".format(output_path))
def main_predict_multitask(input_path, output_path, resources_path, prediction_type, batch_size=64, PADDING_SIZE=30):
"""
:param input_path:
:param output_path:
:param resources_path:
:param prediction_type: either babelnet, wordnet_domains, or lexicographer
:param batch_size: depends on the model
:param PADDING_SIZE: depends on the model
:return: None
"""
K.backend.clear_session()
# ##################
# # vocab loading #
# #################
mapping = pd.read_csv(os.path.join(resources_path, "mapping.csv"))
senses = utils.json_vocab_reader(os.path.join(resources_path, 'semcor.vocab.WordNet.json'))
wordnet_domains_vocabulary = utils.json_vocab_reader(os.path.join(resources_path, 'semcor.vocab.WordNetDomain.json'))
lexicographer_vocabulary = utils.json_vocab_reader(os.path.join(resources_path, 'semcor.vocab.LexNames.json'))
inputs, antivocab = utils.json_vocab_reader(os.path.join(resources_path, 'semcor.input.vocab.json'),
os.path.join(resources_path, 'semcor.leftout.vocab.json'))
output_vocab = utils.merge_vocabulary(senses, inputs)
output_vocab2 = utils.merge_vocabulary(wordnet_domains_vocabulary, inputs)
output_vocab3 = utils.merge_vocabulary(lexicographer_vocabulary, inputs)
reverse_output1_vocab = dict((v, k) for k, v in output_vocab.items())
reverse_output2_vocab = dict((v, k) for k, v in output_vocab2.items())
reverse_output3_vocab = dict((v, k) for k, v in output_vocab3.items())
# ##################
# # Model loading #
# #################
model_path = os.path.join(resources_path, 'models/best_model/model_2019-09-12_12:13:22_+0000.h5')
model_weight_path = os.path.join(resources_path, 'models/best_model/model_weights_2019-09-12_12:13:22_+0000.h5')
loaded_model = K.models.load_model(model_path)
loaded_model.load_weights(model_weight_path)
### generator
eval_generator = generatorMultitask.get(batch_size = 64,
resources_path = resources_path,
training_file_path = input_path,
antivocab = antivocab,
output_vocab = output_vocab,
output_vocab2 = output_vocab2,
output_vocab3 = output_vocab3,
PADDING_SIZE = PADDING_SIZE)
real_words = eval_parser(path = input_path, batch_size = batch_size)
with open(output_path, mode="a") as out:
for batch_ground_truth_sentences in tqdm(real_words, desc='batch: '):
batch_x, candidate_synsets_wordnet, candidates_wndomain, candidates_lex = next(eval_generator)
batch_model_predictions = loaded_model.predict_on_batch(batch_x)
if prediction_type =='babelnet': #actually using babelnet
batch_model_predictions = batch_model_predictions[0]
reverse_output_vocab = reverse_output1_vocab
candidate_synsets = candidate_synsets_wordnet
elif prediction_type =='wordnet_domains':
batch_model_predictions = batch_model_predictions[1]
reverse_output_vocab = reverse_output2_vocab
candidate_synsets = candidates_wndomain
elif prediction_type =='lexicographer':
batch_model_predictions = batch_model_predictions[2]
reverse_output_vocab = reverse_output3_vocab
candidate_synsets = candidates_lex
batch_outputs = basic_predict(batch_ground_truth_sentences,
batch_model_predictions,
candidate_synsets,
PADDING_SIZE,
reverse_output_vocab)
for line in batch_outputs:
if not line.success:
pred = mapping[mapping.WordNet==line.WordNet][prediction_type].values
assert len(pred)==1, "error in mapping {}" .format(line.WordNet)
fmt = "{} {} \n".format(line.Sentence_id, pred[0])
else:
fmt = "{} {} \n".format(line.Sentence_id, line.WordNet) #not really wordnet in this case
out.write(fmt)
print("done writing to:\t{}".format(output_path))