criterion = LabelSmoothing(size=V, padding_idx=0, smoothing=0.0)
criterion.cuda()
model = make_model(V, V, N=transformer_size)
model.cuda()
model_opt = NoamOpt(model.src_embed[0].d_model, 1, 400,
torch.optim.Adam(model.parameters(), lr=0, betas=(0.9, 0.98), eps=1e-9))
for epoch in range(epoch_number):
if epoch % epoches_of_loss_record == 0:
f = open("procedure.txt", "a+")
f.write("step:%d \n" % epoch)
f.close()
print("step: ", epoch)
model.train()
run_epoch("train", data_gen_char(dataloader, batch, nbatches), model,
SimpleLossCompute(model.generator, criterion, model_opt), nbatches, epoch)
model.eval()
run_epoch("test ", data_gen_char(dataloader, batch, 1), model,
SimpleLossCompute(model.generator, criterion, None), nbatches, epoch)
else:
dataloader = DataLoader_token_kg(filename, ents, chunk_len, device)
V = dataloader.vocabularyLoader.n_tokens # vocabolary size
dataloader_ast = DataLoader_token_ast(filename, ents, "Path.txt", chunk_len, device, 9)
V_ast = dataloader_ast.vocabularyLoader.n_token_ast
criterion = LabelSmoothing(size=V, padding_idx=0, smoothing=0.0)
criterion.to(device)
#model = make_model_kg(V, V, "kg_embed/embedding.vec.json", N=transformer_size)
model = make_model_ast(V_ast, V, V, "kg_embed/embedding.vec.json", N=transformer_size)
torch.optim.Adam(model.parameters(),
lr=0,
betas=(0.9, 0.98),
eps=1e-9))
for epoch in range(epoch_number):
if epoch % epoches_of_loss_record == 0:
f = open("procedure.txt", "a+")
f.write("step:%d \n" % epoch)
f.close()
print("step: ", epoch)
model.train()
run_epoch(
"train", data_gen_overlap(dataloader, batch,
nbatches), model,
SimpleLossCompute(model.generator, criterion, model_opt),
nbatches, epoch)
model.eval()
run_epoch("test ", data_gen_overlap(dataloader, batch,
nbatches), model,
SimpleLossCompute(model.generator, criterion, None),
nbatches, epoch)
else:
dataloader = DataLoader_token(filename, chunk_len, device)
V = dataloader.vocabularyLoader.n_tokens # vocabolary size
criterion = LabelSmoothing(size=V, padding_idx=0, smoothing=0.0)
criterion.cuda()
model = make_model(V, V, N=transformer_size)
model.cuda()