def sample(model, data, dev, nsamples=100, use_max=False):
while True:
cmd = input('Do you want to sample text from the model [y/n]')
if cmd == 'n':
return
else:
seed = input('Please input some seeding text, e.g., #include <c: ')
inputs = []
for c in seed:
x = np.zeros((1, data.vocab_size), dtype=np.float32)
x[0, data.char_to_idx[c]] = 1
tx = tensor.from_numpy(x)
tx.to_device(dev)
inputs.append(tx)
model.reset_states(dev)
outputs = model(inputs)
y = tensor.softmax(outputs[-1])
sys.stdout.write(seed)
for i in range(nsamples):
prob = tensor.to_numpy(y)[0]
if use_max:
cur = np.argmax(prob)
else:
cur = np.random.choice(data.vocab_size, 1, p=prob)[0]
sys.stdout.write(data.idx_to_char[cur])
x = np.zeros((1, data.vocab_size), dtype=np.float32)
x[0, cur] = 1
tx = tensor.from_numpy(x)
tx.to_device(dev)
outputs = model([tx])
y = tensor.softmax(outputs[-1])
def predict(self, queries: List[str]):
print("Get queries")
res = []
queries = [self._preprocess(ele) for ele in queries]
for input_ids in queries:
x = tensor.Tensor(device=self.dev, data=input_ids)
out = []
for i in range(self.length):
y = self._model.forward(x)
y = autograd.reshape(y, y.shape[-2:])[-1, :]
y = tensor.softmax(y)
y = tensor.to_numpy(y)[0]
y = np.argsort(y)[-1]
out.append(y)
y = np.array([y]).reshape([1, 1, -1]).astype(np.float32)
y = tensor.Tensor(device=self.dev, data=y)
x = tensor.concatenate([x, y], 2)
result = self._postprocess(out)
res.append(result)
return res
img = preprocess(img)
# sg_ir = sonnx.prepare(onnx_model) # run without graph
# y = sg_ir.run([img])
logging.info("model compling...")
dev = device.create_cuda_gpu()
x = tensor.PlaceHolder(img.shape, device=dev)
model = MyModel(onnx_model)
model.compile([x], is_train=False, use_graph=True, sequential=True)
# verifty the test
# from utils import load_dataset
# inputs, ref_outputs = load_dataset(os.path.join('/tmp', 'model', 'test_shufflenetv2',
# 'model.onnx'))
# x_batch = tensor.Tensor(device=dev, data=inputs[0])
# outputs = sg_ir.run([x_batch])
# for ref_o, o in zip(ref_outputs, outputs):
# np.testing.assert_almost_equal(ref_o, tensor.to_numpy(o), 4)
logging.info("model running...")
x = tensor.Tensor(device=dev, data=img)
y = model.forward(x)
logging.info("postprocessing...")
y = tensor.softmax(y)
scores = tensor.to_numpy(y)
scores = np.squeeze(scores)
a = np.argsort(scores)[::-1]
for i in a[0:5]:
logging.info('class=%s ; probability=%f' % (labels[i], scores[i]))
def sample(model_path, nsamples=100, seed_text='', do_sample=True):
with open(model_path, 'rb') as fd:
d = pickle.load(fd)
rnn_w = tensor.from_numpy(d['rnn_w'])
idx_to_char = d['idx_to_char']
char_to_idx = d['char_to_idx']
vocab_size = len(idx_to_char)
dense_w = tensor.from_numpy(d['dense_w'])
dense_b = tensor.from_numpy(d['dense_b'])
hidden_size = d['hidden_size']
num_stacks = d['num_stacks']
dropout = d['dropout']
cuda = device.create_cuda_gpu()
rnn = layer.LSTM(name='lstm', hidden_size=hidden_size,
num_stacks=num_stacks, dropout=dropout,
input_sample_shape=(len(idx_to_char),))
rnn.to_device(cuda)
rnn.param_values()[0].copy_data(rnn_w)
dense = layer.Dense('dense', vocab_size, input_sample_shape=(hidden_size,))
dense.to_device(cuda)
dense.param_values()[0].copy_data(dense_w)
dense.param_values()[1].copy_data(dense_b)
hx = tensor.Tensor((num_stacks, 1, hidden_size), cuda)
cx = tensor.Tensor((num_stacks, 1, hidden_size), cuda)
hx.set_value(0.0)
cx.set_value(0.0)
if len(seed_text) > 0:
for c in seed_text:
x = np.zeros((1, vocab_size), dtype=np.float32)
x[0, char_to_idx[c]] = 1
tx = tensor.from_numpy(x)
tx.to_device(cuda)
inputs = [tx, hx, cx]
outputs = rnn.forward(False, inputs)
y = dense.forward(False, outputs[0])
y = tensor.softmax(y)
hx = outputs[1]
cx = outputs[2]
sys.stdout.write(seed_text)
else:
y = tensor.Tensor((1, vocab_size), cuda)
y.set_value(1.0 / vocab_size)
for i in range(nsamples):
y.to_host()
prob = tensor.to_numpy(y)[0]
if do_sample:
cur = np.random.choice(vocab_size, 1, p=prob)[0]
else:
cur = np.argmax(prob)
sys.stdout.write(idx_to_char[cur])
x = np.zeros((1, vocab_size), dtype=np.float32)
x[0, cur] = 1
tx = tensor.from_numpy(x)
tx.to_device(cuda)
inputs = [tx, hx, cx]
outputs = rnn.forward(False, inputs)
y = dense.forward(False, outputs[0])
y = tensor.softmax(y)
hx = outputs[1]
cx = outputs[2]
print('')
def sample(model_path, nsamples=100, seed_text='', do_sample=True):
with open(model_path, 'rb') as fd:
d = pickle.load(fd)
rnn_w = tensor.from_numpy(d['rnn_w'])
idx_to_char = d['idx_to_char']
char_to_idx = d['char_to_idx']
vocab_size = len(idx_to_char)
dense_w = tensor.from_numpy(d['dense_w'])
dense_b = tensor.from_numpy(d['dense_b'])
hidden_size = d['hidden_size']
num_stacks = d['num_stacks']
dropout = d['dropout']
cuda = device.create_cuda_gpu()
rnn = layer.LSTM(name='lstm',
hidden_size=hidden_size,
num_stacks=num_stacks,
dropout=dropout,
input_sample_shape=(len(idx_to_char), ))
rnn.to_device(cuda)
rnn.param_values()[0].copy_data(rnn_w)
dense = layer.Dense('dense',
vocab_size,
input_sample_shape=(hidden_size, ))
dense.to_device(cuda)
dense.param_values()[0].copy_data(dense_w)
dense.param_values()[1].copy_data(dense_b)
hx = tensor.Tensor((num_stacks, 1, hidden_size), cuda)
cx = tensor.Tensor((num_stacks, 1, hidden_size), cuda)
hx.set_value(0.0)
cx.set_value(0.0)
if len(seed_text) > 0:
for c in seed_text:
x = np.zeros((1, vocab_size), dtype=np.float32)
x[0, char_to_idx[c]] = 1
tx = tensor.from_numpy(x)
tx.to_device(cuda)
inputs = [tx, hx, cx]
outputs = rnn.forward(False, inputs)
y = dense.forward(False, outputs[0])
y = tensor.softmax(y)
hx = outputs[1]
cx = outputs[2]
sys.stdout.write(seed_text)
else:
y = tensor.Tensor((1, vocab_size), cuda)
y.set_value(1.0 / vocab_size)
for i in range(nsamples):
y.to_host()
prob = tensor.to_numpy(y)[0]
if do_sample:
cur = np.random.choice(vocab_size, 1, p=prob)[0]
else:
cur = np.argmax(prob)
sys.stdout.write(idx_to_char[cur])
x = np.zeros((1, vocab_size), dtype=np.float32)
x[0, cur] = 1
tx = tensor.from_numpy(x)
tx.to_device(cuda)
inputs = [tx, hx, cx]
outputs = rnn.forward(False, inputs)
y = dense.forward(False, outputs[0])
y = tensor.softmax(y)
hx = outputs[1]
cx = outputs[2]
print('')
