def run(p="../../../data/atis/atis.pkl",
wordembdim=70,
lablembdim=70,
innerdim=300,
lr=0.05,
numbats=100,
epochs=20,
validinter=1,
wreg=0.0003,
depth=1):
train, test, dics = pickle.load(open(p))
word2idx = dics["words2idx"]
table2idx = dics["tables2idx"]
label2idx = dics["labels2idx"]
label2idxrev = {v: k for k, v in label2idx.items()}
train = zip(*train)
test = zip(*test)
print "%d training examples, %d test examples" % (len(train), len(test))
#tup2text(train[0], word2idx, table2idx, label2idx)
maxlen = 0
for tup in train + test:
maxlen = max(len(tup[0]), maxlen)
numwords = max(word2idx.values()) + 2
numlabels = max(label2idx.values()) + 2
# get training data
traindata = getdatamatrix(train, maxlen, 0).astype("int32")
traingold = getdatamatrix(train, maxlen, 2).astype("int32")
trainmask = (traindata > 0).astype("float32")
# test data
testdata = getdatamatrix(test, maxlen, 0).astype("int32")
testgold = getdatamatrix(test, maxlen, 2).astype("int32")
testmask = (testdata > 0).astype("float32")
res = atiseval(testgold - 1, testgold - 1, label2idxrev)
print res #; exit()
# define model
innerdim = [innerdim] * depth
m = SimpleSeqTransDec(indim=numwords,
inpembdim=wordembdim,
outembdim=lablembdim,
innerdim=innerdim,
outdim=numlabels)
# training
m = m.train([traindata, shiftdata(traingold), trainmask], traingold).adagrad(lr=lr).grad_total_norm(5.0).seq_cross_entropy().l2(wreg)\
.split_validate(splits=5, random=True).seq_cross_entropy().seq_accuracy().validinter(validinter).takebest()\
.train(numbats, epochs)
# predict after training
s = GreedySearch(m, startsymbol=0)
testpred, _ = s.decode(testdata)
testpred = testpred * testmask
evalres = atiseval(testpred - 1, testgold - 1, label2idxrev)
print evalres
def run(p="../../../data/atis/atis.pkl", wordembdim=70, lablembdim=70, innerdim=300, lr=0.05, numbats=100, epochs=20, validinter=1, wreg=0.0003, depth=1):
train, test, dics = pickle.load(open(p))
word2idx = dics["words2idx"]
table2idx = dics["tables2idx"]
label2idx = dics["labels2idx"]
label2idxrev = {v: k for k, v in label2idx.items()}
train = zip(*train)
test = zip(*test)
print "%d training examples, %d test examples" % (len(train), len(test))
#tup2text(train[0], word2idx, table2idx, label2idx)
maxlen = 0
for tup in train + test:
maxlen = max(len(tup[0]), maxlen)
numwords = max(word2idx.values()) + 2
numlabels = max(label2idx.values()) + 2
# get training data
traindata = getdatamatrix(train, maxlen, 0).astype("int32")
traingold = getdatamatrix(train, maxlen, 2).astype("int32")
trainmask = (traindata > 0).astype("float32")
# test data
testdata = getdatamatrix(test, maxlen, 0).astype("int32")
testgold = getdatamatrix(test, maxlen, 2).astype("int32")
testmask = (testdata > 0).astype("float32")
res = atiseval(testgold-1, testgold-1, label2idxrev); print res#; exit()
# define model
innerdim = [innerdim] * depth
m = SimpleSeqTransDec(indim=numwords, inpembdim=wordembdim, outembdim=lablembdim, innerdim=innerdim, outdim=numlabels)
# training
m = m.train([traindata, shiftdata(traingold), trainmask], traingold).adagrad(lr=lr).grad_total_norm(5.0).seq_cross_entropy().l2(wreg)\
.split_validate(splits=5, random=True).seq_cross_entropy().seq_accuracy().validinter(validinter).takebest()\
.train(numbats, epochs)
# predict after training
s = SeqTransDecSearch(m)
testpred, _ = s.decode(testdata)
testpred = testpred * testmask
evalres = atiseval(testpred-1, testgold-1, label2idxrev); print evalres
def test_output_shape(self):
# settings
batsize = 10
seqlen = 5
invocsize = 50
inembdim = 50
outembdim = 40
innerdim = 11
outvocsize = 17
# data
traindata = np.random.randint(0, invocsize, (batsize, seqlen))
traingold = np.random.randint(0, outvocsize, (batsize, seqlen))
# model
m = SimpleSeqTransDec(indim=invocsize, inpembdim=inembdim, outpembdim=outembdim, innerdim=innerdim, outdim=outvocsize)
pred = m.predict(traindata, shiftdata(traingold))
self.assertEqual(pred.shape, (batsize, seqlen, outvocsize))
def test_output_shape(self):
# settings
batsize = 10
seqlen = 5
invocsize = 50
inembdim = 50
outembdim = 40
innerdim = 11
outvocsize = 17
# data
traindata = np.random.randint(0, invocsize, (batsize, seqlen))
traingold = np.random.randint(0, outvocsize, (batsize, seqlen))
# model
m = SimpleSeqTransDec(
indim=invocsize, inpembdim=inembdim, outpembdim=outembdim, innerdim=innerdim, outdim=outvocsize
)
pred = m.predict(traindata, shiftdata(traingold))
self.assertEqual(pred.shape, (batsize, seqlen, outvocsize))
def test_stop_symbol(self):
m = SimpleSeqTransDec(indim=20,
outdim=10,
inpembdim=8,
outembdim=9,
innerdim=11)
inpseq = np.random.randint(0, 20, (5, 7)).astype("int32")
searcher = GreedySearch(m, startsymbol=0, stopsymbol=0)
searcher.init(5)
out = searcher.search(inpseq)
self.assertTrue(np.all(out[0] == np.zeros_like(out[0])))
def test_recappl_shapes_model_user(self):
batsize = 100
model = SimpleSeqTransDec(indim=200,
outdim=50,
inpembdim=20,
outembdim=20,
innerdim=[40, 30])
mu = RecPredictor(model).init(batsize)
inpval2 = np.random.randint(0, 50, (batsize, )).astype("int32")
for i in range(5):
inpval = np.random.randint(0, 200, (batsize, )).astype("int32")
outpval = mu.feed(inpval, inpval2)
inpval2 = np.argmax(outpval, axis=1).astype("int32")
self.assertEqual(outpval.shape, (batsize, 50))
