def test_run_tf_model(self):
model = os.path.join(MODEL_DIR, 'graph.pb')
bad_model = os.path.join(MODEL_DIR, 'pt-minimal.pt')
with open(model, 'rb') as f:
model_pb = f.read()
with open(bad_model, 'rb') as f:
wrong_model_pb = f.read()
con = self.get_client()
con.modelset('m', Backend.tf, Device.cpu, model_pb,
input=['a', 'b'], output='mul')
# wrong model
self.assertRaises(ResponseError,
con.modelset, 'm', Backend.tf, Device.cpu,
wrong_model_pb,
input=['a', 'b'], output='mul')
# missing inputs/outputs
self.assertRaises(ValueError,
con.modelset, 'm', Backend.tf, Device.cpu,
wrong_model_pb)
# wrong backend
self.assertRaises(ResponseError,
con.modelset, 'm', Backend.torch, Device.cpu,
model_pb,
input=['a', 'b'], output='mul')
con.tensorset('a', Tensor.scalar(DType.float, 2, 3))
con.tensorset('b', Tensor.scalar(DType.float, 2, 3))
con.modelrun('m', ['a', 'b'], 'c')
tensor = con.tensorget('c')
self.assertEqual([4, 9], tensor.value)
def testScriptLoad(self):
con = self.get_client()
dirname = os.path.dirname(__file__)
path = f'{dirname}/testdata/script.txt'
script = raimodel.Model.load(path)
con.scriptset('script', Device.cpu, script)
con.tensorset('a', Tensor.scalar(DType.float, 2, 5))
con.tensorset('b', Tensor.scalar(DType.float, 3, 7))
con.scriptrun('script', 'bar', ['a', 'b'], 'c')
tensor = con.tensorget('c')
self.assertEqual([5, 12], tensor.value)
def test_set_tensor(self):
con = self.get_client()
con.tensorset('x', Tensor.scalar(DType.float, 2, 3))
values = con.tensorget('x')
self.assertEqual([2, 3], values.value)
con.tensorset('x', Tensor.scalar(DType.int32, 2, 3))
values = con.tensorget('x').value
self.assertEqual([2, 3], values)
self.assertRaises(Exception, con.tensorset, 1)
self.assertRaises(Exception, con.tensorset, 'x')
def testPyTorchGraph(self):
torch_graph = MyModule()
path = f'{time.time()}.pb'
raimodel.Model.save(torch_graph, path)
model = raimodel.Model.load(path)
os.remove(path)
con = self.get_client()
con.modelset('ptmodel', Backend.torch, Device.cpu, model)
con.tensorset('a', Tensor.scalar(DType.float, 2, 5))
con.tensorset('b', Tensor.scalar(DType.float, 3, 7))
con.modelrun('ptmodel', ['a', 'b'], 'c')
tensor = con.tensorget('c')
self.assertEqual([5, 12], tensor.value)
def test_scripts(self):
con = self.get_client()
self.assertRaises(ResponseError, con.scriptset,
'ket', Device.cpu, 'return 1')
con.scriptset('ket', Device.cpu, r"""
def bar(a, b):
return a + b
""")
con.tensorset('a', Tensor.scalar(DType.float, 2, 3))
con.tensorset('b', Tensor.scalar(DType.float, 2, 3))
# try with bad arguments:
self.assertRaises(ResponseError,
con.scriptrun, 'ket', 'bar', input='a', output='c')
con.scriptrun('ket', 'bar', input=['a', 'b'], output='c')
tensor = con.tensorget('c')
self.assertEqual([4, 6], tensor.value)
# TODO: image/blob tests; more numpy tests..
def testTFGraph(self):
y = get_tf_graph()
init = tf.global_variables_initializer()
sess = tf.Session()
sess.run(init)
path = f'{time.time()}.pb'
raimodel.Model.save(sess, path, output=['output'])
model = raimodel.Model.load(path)
os.remove(path)
con = self.get_client()
con.modelset('tfmodel',
Backend.tf,
Device.cpu,
model,
input=['input'],
output=['output'])
con.tensorset('a', Tensor.scalar(DType.float, 2))
con.modelrun('tfmodel', ['a'], 'c')
tensor = con.tensorget('c')
self.assertEqual([13], tensor.value)
def testTensorShapes(self):
t = Tensor(DType.float, [4], [1, 2, 3, 4])
self.assertEqual([4], t.shape)
t = BlobTensor.from_numpy(np.array([[1, 2, 3], [4, 5, 6]]))
self.assertEqual([2, 3], t.shape)
from __future__ import print_function
from redisai import Client, Tensor, ScalarTensor, \
BlobTensor, DType, Device, Backend
from redisai import model as raimodel
client = Client()
client.tensorset('x', Tensor(DType.float, [2], [2, 3]))
t = client.tensorget('x')
print(t.value)
model = raimodel.Model.load('../RedisAI/examples/models/graph.pb')
client.tensorset('a', ScalarTensor(DType.float, 2, 3))
client.tensorset('b', ScalarTensor(DType.float, 12, 10))
client.modelset('m', Backend.tf,
Device.cpu,
input=['a', 'b'],
output='mul',
data=model)
client.modelrun('m', ['a', 'b'], ['mul'])
print(client.tensorget('mul').value)
# Try with a script
script = raimodel.Model.load('../RedisAI/examples/models/script.txt')
client.scriptset('ket', Device.cpu, script)
client.scriptrun('ket', 'bar', input=['a', 'b'], output='c')
b1 = client.tensorget('c', astype=BlobTensor)
b2 = client.tensorget('c', astype=BlobTensor)
bt = BlobTensor(DType.float, b1.shape, b1, b2)
print(len(bytes(bt.blob)))