def __init__(self, packer, l_map, SCRF_l_map):
self.packer = packer
self.l_map = l_map
self.SCRF_l_map = SCRF_l_map
self.r_l_map = utils.revlut(l_map)
self.SCRF_r_l_map = utils.revlut(SCRF_l_map)
def __init__(self, packer, l_map, SCRF_l_map):
self.packer = packer
self.l_map = l_map
self.SCRF_l_map = SCRF_l_map
self.r_l_map = utils.revlut(l_map)
self.SCRF_r_l_map = utils.revlut(SCRF_l_map)
def __init__(self, packer, l_map, SCRF_l_map):
self.packer = packer
self.l_map = l_map
self.SCRF_l_map = SCRF_l_map
self.r_l_map = utils.revlut(l_map)
self.SCRF_r_l_map = utils.revlut(SCRF_l_map)
self.true_labels = []
self.scrf_pred_labels = []
self.crf_pred_labels = []
self.joint_pred_labels = []
ts, f2i, _ = conll_load_sentences(args.train, word_vocab, char_vocab, maxs,
maxw, f2i, long_0_tensor_alloc)
print('Loaded training data')
if args.valid is not None:
print('Using provided validation data')
vs, f2i, _ = conll_load_sentences(args.valid, word_vocab, char_vocab, maxs,
maxw, f2i, long_0_tensor_alloc)
else:
ts, vs = valid_split(ts, args.valsplit)
print('Created validation split')
es, f2i, txts = conll_load_sentences(args.test, word_vocab, char_vocab, maxs,
maxw, f2i, long_0_tensor_alloc)
print('Loaded test data')
i2f = revlut(f2i)
print(i2f)
print('Using %d examples for training' % len(ts))
print('Using %d examples for validation' % len(vs))
print('Using %d examples for test' % len(es))
model = TaggerModel(f2i, word_vec, char_vec, maxs, maxw, args.rnn, args.wsz,
args.hsz, args.cfiltsz, args.dropout, args.numrnn)
trainer = Trainer(gpu, model, args.optim, args.eta, args.mom)
outname = '%s/%s.model' % (args.outdir, args.save)
max_acc = 0
last_improved = 0
for i in range(args.epochs):
print('Training epoch %d' % (i + 1))
trainer.train(ts, args.batchsz)
embed1 = Word2VecModel(args.embed1, vocab1, args.unif)
print('Loaded word embeddings: ' + args.embed1)
if args.embed2 is None:
print('No embed2 found, using embed1 for both')
args.embed2 = args.embed1
embed2 = Word2VecModel(args.embed2, vocab2, args.unif)
print('Loaded word embeddings: ' + args.embed2)
ts = load_sentences(args.train, embed1.vocab, embed2.vocab, args.mxlen,
args.batchsz, long_0_tensor_alloc)
es = load_sentences(args.test, embed1.vocab, embed2.vocab, args.mxlen,
args.batchsz, long_0_tensor_alloc)
rlut1 = revlut(embed1.vocab)
rlut2 = revlut(embed2.vocab)
Seq2SeqModelType = Seq2SeqAttnModel if args.attn else Seq2SeqModel
print(Seq2SeqModelType)
seq2seq = Seq2SeqModelType(embed1, embed2, args.mxlen, args.hsz, args.layers,
args.rnntype)
trainer = Trainer(gpu, seq2seq, args.optim, args.eta, args.mom)
err_min = 1
last_improved = 0
reset = 0
for i in range(args.epochs):
print('Training epoch %d' % (i + 1))
def __init__(self, packer, l_map):
self.packer = packer
self.l_map = l_map
self.r_l_map = utils.revlut(l_map)
f2i,
args.clean,
args.chars,
args.mxlen,
mxfiltsz,
vec_alloc=long_0_tensor_alloc,
ExType=TorchExamples)
print('Loaded test data')
nc = len(f2i)
mdsave(f2i, embeddings.vocab, args.outdir, args.save)
model = ConvModel(embeddings, nc, args.filtsz, args.cmotsz, args.hsz,
args.dropout, not args.static)
labels = revlut(f2i)
trainer = Trainer(gpu, model, args.optim, args.eta, args.mom)
max_acc = 0
last_improved = 0
confusion = ConfusionMatrix(labels)
for i in range(args.epochs):
print('Training epoch %d' % (i + 1))
trainer.train(ts, confusion, args.batchsz)
this_acc = trainer.test(vs, confusion, args.batchsz, 'Validation')
if this_acc > max_acc:
max_acc = this_acc
last_improved = i
model.save(args.outdir, args.save)
print('Highest dev acc achieved yet -- writing model')