dropout_p = 0.05
teacher_forcing_ratio = .5
clip = 5.
criterion = nn.NLLLoss()
# Initialize models
encoder = EncoderRNN(input_lang.n_words, hidden_size, n_layers)
decoder = AttentionDecoderRNN(attn_model,
hidden_size,
output_lang.n_words,
n_layers,
dropout_p=dropout_p)
learning_rate = 1
encoder_optimizer = optim.Adam(encoder.parameters(), lr=learning_rate)
decoder_optimizer = optim.Adam(decoder.parameters(), lr=learning_rate)
# Load model parameters
encoder.load_state_dict(torch.load(
'./data/encoder_params_{}'.format(language)))
decoder.load_state_dict(torch.load(
'./data/decoder_params_{}'.format(language)))
decoder.attention.load_state_dict(
torch.load('./data/attention_params_{}'.format(language)))
# Move models to GPU
# encoder.cuda()
# decoder.cuda()
def GetData(input_data):
input_lang, output_lang, pairs = etl.prepare_data(args.language)
# Initialize models
encoder = EncoderRNN(input_lang.n_words, args.embedding_size, args.hidden_size,
args.n_layers, args.dropout)
decoder = AttentionDecoderRNN(output_lang.n_words, args.embedding_size,
args.hidden_size, args.attn_model, args.n_layers,
args.dropout)
# Move models to device
encoder = encoder.to(device)
decoder = decoder.to(device)
# Initialize optimizers and criterion
encoder_optimizer = optim.Adam(encoder.parameters(), lr=args.lr)
decoder_optimizer = optim.Adam(decoder.parameters(), lr=args.lr)
criterion = nn.NLLLoss()
# Keep track of time elapsed and running averages
start = time.time()
plot_losses = []
print_loss_total = 0 # Reset every print_every
plot_loss_total = 0 # Reset every plot_every
# Begin training
for epoch in range(1, args.n_epochs + 1):
# Get training data for this cycle
training_pair = etl.tensor_from_pair(random.choice(pairs), input_lang,
output_lang)
input = training_pair[0]
target = training_pair[1]