def quick_run(cls, runScale='basic', dataScale='tiny_fake_2', useCPU=True):
# ok this is silly. But at least it's fast.
vocabSize = TextDataReader.maker_from_premade_source(dataScale)(
bucketingOrRandom='bucketing', batchSize_=50,
minimumWords=0).vocabSize
params = [('initialLearningRate', [1e-3]), ('l2RegLambda', [0]),
('vocabSize', [vocabSize]), ('embeddingDim', [32]),
('filterSizes', [[2, 4], [1, 3, 5]]),
('numFeaturesPerFilter', [8])]
cls.run_thru_data(TextDataReader, dataScale, make_params_dict(params),
runScale, useCPU)
def full_run(cls, runScale='full', dataScale='full', useCPU=True):
# ok this is silly. But at least it's fast.
vocabSize = TextDataReader.maker_from_premade_source(dataScale)(
bucketingOrRandom='bucketing', batchSize_=50, minimumWords=0).vocabSize
params = [('initialLearningRate', [1e-3]),
('l2RegLambda', [0, 1e-5]),
('vocabSize', [vocabSize]),
('embeddingDim', [64, 128, 300]),
('filterSizes', [[1, 2, 4], [3, 5, 10, 15]]),
('numFeaturesPerFilter', [16, 32, 64]),
('pooledKeepProb', [0.5, 0.7, 0.9, 1])]
cls.run_thru_data(TextDataReader, dataScale, make_params_dict(params), runScale, useCPU)
def quick_run(cls, runScale='basic', dataScale='tiny_fake_2', useCPU=True):
# ok this is silly. But at least it's fast.
vocabSize = TextDataReader.maker_from_premade_source(dataScale)(
bucketingOrRandom='bucketing', batchSize_=50,
minimumWords=0).vocabSize
params = [('initialLearningRate', [1e-3]), ('l2RegLambda', [0]),
('vocabSize', [vocabSize]), ('embeddingDim', [32]),
('numRnnOutputSteps', [10]),
('rnnCellUnitsNProbs', [([3], [0.9]), ([4, 8], [1, 1])]),
('convNumFeaturesPerFilter', [16]), ('pooledKeepProb', [1])]
cls.run_thru_data(TextDataReader, dataScale, make_params_dict(params),
runScale, useCPU)
def quick_learn(cls,
runScale='small',
dataScale='full_2occupations',
useCPU=True):
# ok this is silly. But at least it's fast.
vocabSize = TextDataReader.maker_from_premade_source(dataScale)(
bucketingOrRandom='bucketing', batchSize_=50,
minimumWords=0).vocabSize
params = [('initialLearningRate', [1e-4]), ('l2RegLambda', [1e-5]),
('vocabSize', [vocabSize]), ('embeddingDim', [256]),
('convFilterSizesNKeepProbs', [([2, 3, 5], [0.6, 0.6,
0.6])]),
('numFeaturesPerFilter', [32]), ('pooledKeepProb', [0.9])]
cls.run_thru_data(TextDataReader, dataScale, make_params_dict(params),
runScale, useCPU)
def comparison_run(cls,
runScale='medium',
dataScale='full_2occupations',
useCPU=True):
# ok this is silly. But at least it's fast.
vocabSize = TextDataReader.maker_from_premade_source(dataScale)(
bucketingOrRandom='bucketing', batchSize_=50,
minimumWords=0).vocabSize
params = [('initialLearningRate', [1e-3]), ('l2RegLambda', [5e-4]),
('vocabSize', [vocabSize]), ('embeddingDim', [128, 300]),
('numRnnOutputSteps', [5, 10]),
('rnnCellUnitsNProbs', [([64, 64, 32], [0.8, 0.8, 0.9])]),
('convNumFeaturesPerFilter', [16]),
('pooledKeepProb', [0.5, 0.9])]
cls.run_thru_data(TextDataReader, dataScale, make_params_dict(params),
runScale, useCPU)
[self.vocabSize, self.embeddingDim], -1.0, 1.0),
name="W")
self.output = tf.nn.embedding_lookup(self.W, self.input)
@property
def output_shape(self):
return self.inputDim[0], self.inputDim[1], self.embeddingDim
@classmethod
def new(cls, vocabSize_, embeddingDim, activation=None):
return lambda input_, inputDim_, loggerFactory=None: \
cls(input_, inputDim_, vocabSize_, embeddingDim, activation, loggerFactory)
if __name__ == '__main__':
dr = TextDataReader('../data/peopleData/tokensfiles/pol_sci.json',
'bucketing', 5, 1)
fd = dr.get_next_training_batch()[0]
# inputVal = np.array([[4, 5, 1, 0, 0], [2, 0, 0, 0, 0]])
# inputShape = inputVal.shape
# v = tf.Variable(inputVal)
maker = EmbeddingLayer.new(vocabSize_=dr.vocabSize, embeddingDim=32)
layer = maker(dr.input['x'], [-1, dr.maxXLen])
sess = tf.InteractiveSession()
sess.run(tf.global_variables_initializer())
# output = sess.run(layer.output)
output = sess.run(layer.output, fd)
print('-------- INPUT --------')