print "Dev examples " + str(devCounter)
devSurprisalTableHere = [
surp / (devCounter * args.batchSize) for surp in surprisalTable
]
return devLoss / devWords, devSurprisalTableHere, devMemory / devCounter
wordsProcessed = 0
startTime = time.time()
DEV_PERIOD = 10000
epochCount = 0
while failedDevRuns < 10:
epochCount += 1
print "Starting new epoch, permuting corpus"
corpus = corpusIteratorWikiWords.training(args.language)
# stream = createStream(corpus)
stream = prepareDatasetChunks(corpus, train=True)
while True:
counter += 1
printHere = (counter % 50 == 0)
if counter % DEV_PERIOD == 0:
hidden = None
beginning = zeroBeginning
newDevLoss, devSurprisalTableHere, newDevMemory = computeDevLoss()
hidden = None
beginning = zeroBeginning
loss.backward()
torch.nn.utils.clip_grad_value_(parameters_cached,
5.0) #, norm_type="inf")
optim.step()
lossHasBeenBad = 0
import time
totalStartTime = time.time()
devLosses = []
for epoch in range(10000):
print(epoch)
training_data_1 = corpusIteratorWikiWords.training(args.language1)
training_data_2 = corpusIteratorWikiWords.training(args.language2)
print("Got data")
training_chars = prepareDatasetChunksTwo(training_data_1,
training_data_2,
train=True)
rnn_drop.train(True)
startTime = time.time()
trainChars = 0
counter = 0
hidden, beginning = None, None
while True:
counter += 1
try:
] #print >> outFile, " ".join(map(str,devLosses))
devSurprisalTable = [
float(x) for x in next(inFile).strip().split(" ")
] # print >> outFile, " ".join(map(str,devSurprisalTable))
devMemories = [float(x) for x in next(inFile).strip().split(" ")
] # print >> outFile, " ".join(map(str, devMemories))
wordsProcessed = 0
startTime = time.time()
DEV_PERIOD = 30000
epochCount = 0
while failedDevRuns < 10:
epochCount += 1
print "Starting new epoch, permuting corpus"
corpus_1 = corpusIteratorWikiWords.training(args.language1)
corpus_2 = corpusIteratorWikiWords.training(args.language2)
# stream = createStream(corpus)
stream = prepareDatasetChunksTwo(corpus_1, corpus_2, train=True)
while True:
counter += 1
printHere = (counter % 50 == 0)
if counter % DEV_PERIOD == 0:
hidden = None
beginning = zeroBeginning
newDevLoss, devSurprisalTableHere, newDevMemory = computeDevLoss()
import corpusIteratorWikiWords
for chunk in corpusIteratorWikiWords.training("english"):
print(chunk)
