def test_evaluate(self):
modelBuilder = PytorchModelBuilder(model_creator=model_creator_pytorch,
optimizer_creator=optimizer_creator,
loss_creator=loss_creator)
model = modelBuilder.build(config={
"lr": 1e-2,
"batch_size": 32,
})
model.fit_eval(data=(self.data["x"], self.data["y"]),
validation_data=(self.data["val_x"],
self.data["val_y"]),
epochs=20)
mse_eval = model.evaluate(x=self.data["val_x"], y=self.data["val_y"])
try:
import onnx
import onnxruntime
mse_eval_onnx = model.evaluate_with_onnx(x=self.data["val_x"],
y=self.data["val_y"])
np.testing.assert_almost_equal(mse_eval, mse_eval_onnx)
except ImportError:
pass
# incremental training test
model.fit_eval(data=(self.data["x"], self.data["y"]),
validation_data=(self.data["val_x"],
self.data["val_y"]),
epochs=20)
mse_eval = model.evaluate(x=self.data["val_x"], y=self.data["val_y"])
try:
import onnx
import onnxruntime
mse_eval_onnx = model.evaluate_with_onnx(x=self.data["val_x"],
y=self.data["val_y"])
np.testing.assert_almost_equal(mse_eval, mse_eval_onnx)
except ImportError:
pass
def test_create_not_torch_model(self):
def model_creator(config):
return torch.Tensor(3, 5)
modelBuilder = PytorchModelBuilder(model_creator=model_creator,
optimizer_creator=optimizer_creator,
loss_creator=loss_creator)
with pytest.raises(ValueError):
model = modelBuilder.build(config={
"lr": 1e-2,
"batch_size": 32,
})
def test_fit_evaluate(self):
modelBuilder = PytorchModelBuilder(model_creator=model_creator_pytorch,
optimizer_creator=optimizer_creator,
loss_creator=loss_creator)
model = modelBuilder.build(config={
"lr": 1e-2,
"batch_size": 32,
})
val_result = model.fit_eval(data=(self.data["x"], self.data["y"]),
validation_data=(self.data["val_x"],
self.data["val_y"]),
epochs=20)
assert val_result is not None
def test_dataloader_fit_evaluate(self):
modelBuilder = PytorchModelBuilder(model_creator=model_creator_pytorch,
optimizer_creator=optimizer_creator,
loss_creator=loss_creator)
model = modelBuilder.build(
config={
"lr": 1e-2,
"batch_size": 32,
"train_size": 500,
"valid_size": 100,
"shuffle": True
})
val_result = model.fit_eval(data=train_dataloader_creator,
validation_data=valid_dataloader_creator,
epochs=20)
assert model.config["train_size"] == 500
assert model.config["valid_size"] == 100
assert model.config["shuffle"] is True
assert val_result is not None
def test_predict(self):
modelBuilder = PytorchModelBuilder(model_creator=model_creator_pytorch,
optimizer_creator=optimizer_creator,
loss_creator=loss_creator)
model = modelBuilder.build(config={
"lr": 1e-2,
"batch_size": 32,
})
model.fit_eval(data=(self.data["x"], self.data["y"]),
validation_data=(self.data["val_x"],
self.data["val_y"]),
epochs=20)
pred = model.predict(x=self.data["val_x"])
pred_full_batch = model.predict(x=self.data["val_x"],
batch_size=len(self.data["val_x"]))
np.testing.assert_almost_equal(pred, pred_full_batch)
try:
import onnx
import onnxruntime
pred_onnx = model.predict_with_onnx(x=self.data["val_x"])
np.testing.assert_almost_equal(pred, pred_onnx)
except ImportError:
pass