public override List <Tensor> Call(List <Array> inputs)
{
var feed_dict = new Dictionary <Variable, Array>();
foreach (var(tensor, value) in Enumerable.Zip(this.placeholders, inputs, (a, b) => (a, b)))
{
Type t = value.GetInnerMostType();
Array v = value;
// cntk only support calculate on float, do auto cast here
if (t != typeof(float) && t != typeof(double))
{
v = MatrixEx.Convert <double>(value);
}
feed_dict[tensor] = v;
}
var updated = new List <Tensor>();
if (this.trainer != null)
{
var input_dict = new UnorderedMapVariableValuePtr();
foreach (Variable argument in this.loss.Arguments)
{
if (feed_dict.ContainsKey(argument))
{
input_dict[argument] = new Value(CNTKBackend.In(feed_dict[argument]));
}
else
{
throw new Exception($"CNTK backend: argument {argument.Name} is not found in inputs. Please double check the model and inputs in 'train_function'.");
}
}
var result = this.trainer.TrainMinibatch(input_dict, this.trainer_output);
foreach (Variable o in this.trainer_output.Keys)
{
updated.Add(c.Out(this.trainer_output[o]));
}
}
if (this.metrics_func != null)
{
var input_dict = new Dictionary <Variable, Value>();
foreach (Variable argument in this.metrics_func.Arguments)
{
if (feed_dict.ContainsKey(argument))
{
input_dict[argument] = new Value(CNTKBackend.In(feed_dict[argument]));
}
else
{
throw new Exception($"CNTK backend: metrics argument {argument.Name} is not found in inputs. Please double check the model and inputs.");
}
}
var output_values = new Dictionary <Variable, Value>();
foreach (Variable variable in this.metrics_outputs)
{
output_values[variable] = null;
}
this.metrics_func.Evaluate(input_dict, output_values, DeviceDescriptor.CPUDevice);
foreach (Variable o in this.metrics_outputs)
{
Value value = output_values[o];
var v = c.Out(value);
updated.Add(v);
}
}
if (this.unrelated_updates != null)
{
var input_dict = new Dictionary <Variable, Value>();
foreach (Variable argument in this.unrelated_updates.Arguments)
{
if (feed_dict.ContainsKey(argument))
{
input_dict[argument] = new Value(CNTKBackend.In(feed_dict[argument]));
}
else
{
throw new Exception($"CNTK backend: assign ops argument {argument.Name} is not found in inputs. Please double check the model and inputs.");
}
}
var output_values = new Dictionary <Variable, Value>();
this.unrelated_updates.Evaluate(input_dict, output_values, DeviceDescriptor.CPUDevice);
}
return(updated);
}
public CNTKFunction(CNTKBackend c, List <Variable> inputs, CNTK.Function[] outputs, List <List <Tensor> > updates, string name)
{
this.c = c;
this.placeholders = inputs;
this.trainer = null;
this.unrelated_updates = null;
this.updates = updates;
if (updates.Count > 0)
{
if (len(outputs) <= 0)
{
throw new Exception();
}
this.loss = outputs[0];
// need group update by gradient place holder
var u_ops = new List <CNTK.Function>();
var unrelated_updates = new List <CNTK.Function>();
foreach (List <Tensor> update in updates)
{
CNTK.Function u;
if (update.Count == 1)
{
u = c.In(update[0]);
}
else if (update.Count == 2)
{
u = C.Assign(c.In(update[0]), c.In(update[1]));
}
else
{
throw new NotImplementedException();
}
if (u.Inputs.Count == 0)
{
u_ops.Add(u);
}
else
{
unrelated_updates.Add(u);
}
}
var update_func = C.Combine(new VariableVector(u_ops.Select(u => u.Output).ToArray()));
CNTK.Function[] grads = update_func.FindAllWithName("keras_grad_placeholder").ToArray();
var u_list = new List <CNTK.Function>();
var p_list = new List <CNTK.Parameter>();
foreach (CNTK.Function g in grads)
{
if (c.grad_parameter_dict.ContainsKey(g))
{
p_list.Add(c.grad_parameter_dict[g]);
u_list.Add(g);
}
else
{
throw new Exception($"CNTK backend: when constructing trainer, found gradient node {g} which is not related to any parameters in the model. Please double check how the gradient node is constructed.");
}
}
if (len(u_list) > 0)
{
Learner learner = Learner.SGDLearner(p_list, new TrainingParameterScheduleDouble(0));
var criterion = (len(outputs) > 1) ?
C.Combine(new VariableVector(new[] { outputs[0], outputs[1] })) :
outputs[0];
this.trainer = Trainer.CreateTrainer(model: outputs[0], lossFunction: criterion, evaluationFunction: null, parameterLearners: new[] { learner });
this.trainer_output = new UnorderedMapVariableValuePtr();
foreach (CNTK.Function f in outputs)
{
this.trainer_output.Add(f, null);
}
}
else if (len(u_ops) > 0)
{
unrelated_updates.AddRange(u_ops);
}
if (len(unrelated_updates) > 0)
{
this.unrelated_updates = C.Combine(new VariableVector(unrelated_updates.Select(_ => _.Output).ToArray()));
}
}
if (this.trainer == null)
{
this.metrics_outputs = outputs.Select(f => f.Output).ToArray();
this.metrics_func = C.Combine(new VariableVector(this.metrics_outputs));
// cntk only could handle loss and 1 metric in trainer, for metrics more
// than 2, need manual eval
}
else if (len(outputs) > 2)
{
this.metrics_outputs = Matrix.Get(outputs, 2, 0).Select(f => f.Output).ToArray();
this.metrics_func = C.Combine(new VariableVector(this.metrics_outputs));
}
else
{
this.metrics_func = null;
}
}
public CNTKFunction(CNTKBackend c, Variable[] inputs, CNTK.Variable[] outputs, List <List <Tensor> > updates, string name)
{
// https://github.com/fchollet/keras/blob/f65a56fb65062c8d14d215c9f4b1015b97cc5bf3/keras/backend/cntk_backend.py#L1501
this.c = c;
this.placeholders = inputs;
this.trainer = null;
this.unrelated_updates = null;
this.updates = updates;
if (updates.Count > 0)
{
if (len(outputs) <= 0)
{
throw new Exception();
}
this.loss = outputs[0];
// need group update by gradient place holder
var u_ops = new List <CNTK.Function>();
var unrelated_updates = new List <CNTK.Function>();
foreach (List <Tensor> update in updates)
{
CNTK.Function u;
if (update.Count == 1)
{
u = c.In(update[0]);
}
else if (update.Count == 2)
{
u = C.Assign(c.In(update[0]), c.In(update[1]));
}
else
{
throw new NotImplementedException();
}
if (u.Arguments.Count == 0)
{
u_ops.Add(u);
}
else
{
unrelated_updates.Add(u);
}
}
var update_func = C.Combine(new VariableVector(u_ops.Select(u => u.Output).ToArray()));
CNTK.Constant[] grads = update_func.Inputs.Where(x => x.Name == "keras_grad_placeholder").Select(x => new Constant(x)).ToArray();
var u_list = new List <CNTK.Constant>();
var p_list = new List <CNTK.Parameter>();
foreach (CNTK.Constant g in grads)
{
if (c.grad_parameter_dict.ContainsKey(g.Uid))
{
p_list.Add(c.grad_parameter_dict[g.Uid]);
u_list.Add(g);
}
else
{
throw new Exception($"CNTK backend: when constructing trainer, found gradient node {g} which is not related to any parameters in the model. Please double check how the gradient node is constructed.");
}
}
if (len(u_list) > 0)
{
Learner learner = Learner.SGDLearner(p_list, new TrainingParameterScheduleDouble(1));
this.trainer = Trainer.CreateTrainer(model: outputs[0],
lossFunction: outputs[0],
evaluationFunction: outputs[1],
parameterLearners: new[] { learner });
}
else if (len(u_ops) > 0)
{
unrelated_updates.AddRange(u_ops);
}
if (len(unrelated_updates) > 0)
{
this.unrelated_updates = C.Combine(new VariableVector(unrelated_updates.Select(_ => _.Output).ToArray()));
}
}
if (this.trainer == null)
{
this.metrics_outputs = outputs;
this.metrics_func = C.Combine(new VariableVector(this.metrics_outputs));
// cntk only could handle loss and 1 metric in trainer, for metrics more
// than 2, need manual eval
}
else if (len(outputs) > 2)
{
this.metrics_outputs = Matrix.Get(outputs, 2, 0);
this.metrics_func = C.Combine(new VariableVector(this.metrics_outputs));
}
else
{
this.metrics_func = null;
}
}