torch-rb 0.4.0 → 0.5.2

data/ext/torch/ext.cpp

@@ -348,16 +348,6 @@ void Init_ext()
       *[](Tensor& self) {
         return self.is_contiguous();
       })
-    .define_method(
-      "addcmul!",
-      *[](Tensor& self, Scalar value, const Tensor & tensor1, const Tensor & tensor2) {
-        return self.addcmul_(tensor1, tensor2, value);
-      })
-    .define_method(
-      "addcdiv!",
-      *[](Tensor& self, Scalar value, const Tensor & tensor1, const Tensor & tensor2) {
-        return self.addcdiv_(tensor1, tensor2, value);
-      })
     .define_method(
       "_requires_grad!",
       *[](Tensor& self, bool requires_grad) {
@@ -372,7 +362,7 @@ void Init_ext()
     .define_method(
       "grad=",
       *[](Tensor& self, torch::Tensor& grad) {
-        self.grad() = grad;
+        self.mutable_grad() = grad;
       })
     .define_method(
       "_dtype",
@@ -609,7 +599,7 @@ void Init_ext()
     .define_method(
       "grad=",
       *[](Parameter& self, torch::Tensor& grad) {
-        self.grad() = grad;
+        self.mutable_grad() = grad;
       });
 
   Class rb_cDevice = define_class_under<torch::Device>(rb_mTorch, "Device")

data/ext/torch/ruby_arg_parser.h

@@ -91,7 +91,7 @@ struct RubyArgs {
   inline c10::optional<int64_t> toInt64Optional(int i);
   inline c10::optional<bool> toBoolOptional(int i);
   inline c10::optional<double> toDoubleOptional(int i);
-  // inline c10::OptionalArray<double> doublelistOptional(int i);
+  inline c10::OptionalArray<double> doublelistOptional(int i);
   // inline at::Layout layout(int i);
   // inline at::Layout layoutWithDefault(int i, at::Layout default_layout);
   inline c10::optional<at::Layout> layoutOptional(int i);
@@ -105,7 +105,7 @@ struct RubyArgs {
   inline c10::optional<at::MemoryFormat> memoryformatOptional(int i);
   // inline at::QScheme toQScheme(int i);
   inline std::string string(int i);
-  // inline c10::optional<std::string> stringOptional(int i);
+  inline c10::optional<std::string> stringOptional(int i);
   // inline PyObject* pyobject(int i);
   inline int64_t toInt64(int i);
   // inline int64_t toInt64WithDefault(int i, int64_t default_int);
@@ -249,6 +249,25 @@ inline c10::optional<double> RubyArgs::toDoubleOptional(int i) {
   return toDouble(i);
 }
 
+inline c10::OptionalArray<double> RubyArgs::doublelistOptional(int i) {
+  if (NIL_P(args[i])) return {};
+
+  VALUE arg = args[i];
+  auto size = RARRAY_LEN(arg);
+  std::vector<double> res(size);
+  for (idx = 0; idx < size; idx++) {
+    VALUE obj = rb_ary_entry(arg, idx);
+    if (FIXNUM_P(obj) || RB_FLOAT_TYPE_P(obj)) {
+      res[idx] = from_ruby<double>(obj);
+    } else {
+      rb_raise(rb_eArgError, "%s(): argument '%s' must be %s, but found element of type %s at pos %d",
+          signature.name.c_str(), signature.params[i].name.c_str(),
+          signature.params[i].type_name().c_str(), rb_obj_classname(obj), idx + 1);
+    }
+  }
+  return res;
+}
+
 inline c10::optional<at::Layout> RubyArgs::layoutOptional(int i) {
   if (NIL_P(args[i])) return c10::nullopt;
 
@@ -285,6 +304,11 @@ inline std::string RubyArgs::string(int i) {
   return from_ruby<std::string>(args[i]);
 }
 
+inline c10::optional<std::string> RubyArgs::stringOptional(int i) {
+  if (!args[i]) return c10::nullopt;
+  return from_ruby<std::string>(args[i]);
+}
+
 inline int64_t RubyArgs::toInt64(int i) {
   if (NIL_P(args[i])) return signature.params[i].default_int;
   return from_ruby<int64_t>(args[i]);

data/ext/torch/templates.h

@@ -19,6 +19,7 @@ using torch::TensorOptions;
 using torch::Layout;
 using torch::MemoryFormat;
 using torch::IntArrayRef;
+using torch::ArrayRef;
 using torch::TensorList;
 using torch::Storage;
 

data/ext/torch/wrap_outputs.h

@@ -90,3 +90,10 @@ inline Object wrap(torch::TensorList x) {
   }
   return Object(a);
 }
+
+inline Object wrap(std::tuple<double, double> x) {
+  Array a;
+  a.push(to_ruby<double>(std::get<0>(x)));
+  a.push(to_ruby<double>(std::get<1>(x)));
+  return Object(a);
+}

data/lib/torch.rb

@@ -261,6 +261,8 @@ module Torch
         Torch._from_blob(bytes, [bytes.bytesize], TensorOptions.new.dtype(DTYPE_TO_ENUM[dtype]))
       elsif args.size == 1 && args.first.is_a?(Array)
         Torch.tensor(args.first, dtype: dtype, device: device)
+      elsif args.size == 0
+        Torch.empty(0, dtype: dtype, device: device)
       else
         Torch.empty(*args, dtype: dtype, device: device)
       end
@@ -434,7 +436,8 @@ module Torch
       zeros(input.size, **like_options(input, options))
     end
 
-    def stft(input, n_fft, hop_length: nil, win_length: nil, window: nil, center: true, pad_mode: "reflect", normalized: false, onesided: true)
+    # center option
+    def stft(input, n_fft, hop_length: nil, win_length: nil, window: nil, center: true, pad_mode: "reflect", normalized: false, onesided: true, return_complex: nil)
       if center
         signal_dim = input.dim
         extended_shape = [1] * (3 - signal_dim) + input.size
@@ -442,12 +445,7 @@ module Torch
         input = NN::F.pad(input.view(extended_shape), [pad, pad], mode: pad_mode)
         input = input.view(input.shape[-signal_dim..-1])
       end
-      _stft(input, n_fft, hop_length, win_length, window, normalized, onesided)
-    end
-
-    def clamp(tensor, min, max)
-      tensor = _clamp_min(tensor, min)
-      _clamp_max(tensor, max)
+      _stft(input, n_fft, hop_length, win_length, window, normalized, onesided, return_complex)
     end
 
     private

data/lib/torch/optim/adadelta.rb

@@ -39,14 +39,14 @@ module Torch
             state[:step] += 1
 
             if group[:weight_decay] != 0
-              grad = grad.add(group[:weight_decay], p.data)
+              grad = grad.add(p.data, alpha: group[:weight_decay])
             end
 
-            square_avg.mul!(rho).addcmul!(1 - rho, grad, grad)
+            square_avg.mul!(rho).addcmul!(grad, grad, value: 1 - rho)
             std = square_avg.add(eps).sqrt!
             delta = acc_delta.add(eps).sqrt!.div!(std).mul!(grad)
             p.data.add!(delta, alpha: -group[:lr])
-            acc_delta.mul!(rho).addcmul!(1 - rho, delta, delta)
+            acc_delta.mul!(rho).addcmul!(delta, delta, value: 1 - rho)
           end
         end
 

data/lib/torch/optim/adagrad.rb

@@ -49,7 +49,7 @@ module Torch
               if p.grad.data.sparse?
                 raise Error, "weight_decay option is not compatible with sparse gradients"
               end
-              grad = grad.add(group[:weight_decay], p.data)
+              grad = grad.add(p.data, alpha: group[:weight_decay])
             end
 
             clr = group[:lr] / (1 + (state[:step] - 1) * group[:lr_decay])
@@ -57,9 +57,9 @@ module Torch
             if grad.sparse?
               raise NotImplementedYet
             else
-              state[:sum].addcmul!(1, grad, grad)
+              state[:sum].addcmul!(grad, grad, value: 1)
               std = state[:sum].sqrt.add!(group[:eps])
-              p.data.addcdiv!(-clr, grad, std)
+              p.data.addcdiv!(grad, std, value: -clr)
             end
           end
         end

data/lib/torch/optim/adam.rb

@@ -58,7 +58,7 @@ module Torch
 
             # Decay the first and second moment running average coefficient
             exp_avg.mul!(beta1).add!(grad, alpha: 1 - beta1)
-            exp_avg_sq.mul!(beta2).addcmul!(1 - beta2, grad, grad)
+            exp_avg_sq.mul!(beta2).addcmul!(grad, grad, value: 1 - beta2)
             if amsgrad
               # Maintains the maximum of all 2nd moment running avg. till now
               Torch.max(max_exp_avg_sq, exp_avg_sq, out: max_exp_avg_sq)
@@ -70,7 +70,7 @@ module Torch
 
             step_size = group[:lr] / bias_correction1
 
-            p.data.addcdiv!(-step_size, exp_avg, denom)
+            p.data.addcdiv!(exp_avg, denom, value: -step_size)
           end
         end
 

data/lib/torch/optim/adamax.rb

@@ -42,7 +42,7 @@ module Torch
             state[:step] += 1
 
             if group[:weight_decay] != 0
-              grad = grad.add(group[:weight_decay], p.data)
+              grad = grad.add(p.data, alpha: group[:weight_decay])
             end
 
             # Update biased first moment estimate.
@@ -57,7 +57,7 @@ module Torch
             bias_correction = 1 - beta1 ** state[:step]
             clr = group[:lr] / bias_correction
 
-            p.data.addcdiv!(-clr, exp_avg, exp_inf)
+            p.data.addcdiv!(exp_avg, exp_inf, value: -clr)
           end
         end
 

data/lib/torch/optim/adamw.rb

@@ -59,7 +59,7 @@ module Torch
 
             # Decay the first and second moment running average coefficient
             exp_avg.mul!(beta1).add!(grad, alpha: 1 - beta1)
-            exp_avg_sq.mul!(beta2).addcmul!(1 - beta2, grad, grad)
+            exp_avg_sq.mul!(beta2).addcmul!(grad, grad, value: 1 - beta2)
             if amsgrad
               # Maintains the maximum of all 2nd moment running avg. till now
               Torch.max(max_exp_avg_sq, exp_avg_sq, out: max_exp_avg_sq)
@@ -71,7 +71,7 @@ module Torch
 
             step_size = group[:lr] / bias_correction1
 
-            p.data.addcdiv!(-step_size, exp_avg, denom)
+            p.data.addcdiv!(exp_avg, denom, value: -step_size)
           end
         end
 

data/lib/torch/optim/asgd.rb

@@ -36,7 +36,7 @@ module Torch
             state[:step] += 1
 
             if group[:weight_decay] != 0
-              grad = grad.add(group[:weight_decay], p.data)
+              grad = grad.add(p.data, alpha: group[:weight_decay])
             end
 
             # decay term

data/lib/torch/optim/rmsprop.rb

@@ -46,25 +46,25 @@ module Torch
             state[:step] += 1
 
             if group[:weight_decay] != 0
-              grad = grad.add(group[:weight_decay], p.data)
+              grad = grad.add(p.data, alpha: group[:weight_decay])
             end
 
-            square_avg.mul!(alpha).addcmul!(1 - alpha, grad, grad)
+            square_avg.mul!(alpha).addcmul!(grad, grad, value: 1 - alpha)
 
             if group[:centered]
               grad_avg = state[:grad_avg]
-              grad_avg.mul!(alpha).add!(1 - alpha, grad)
-              avg = square_avg.addcmul(-1, grad_avg, grad_avg).sqrt!.add!(group[:eps])
+              grad_avg.mul!(alpha).add!(grad, alpha: 1 - alpha)
+              avg = square_avg.addcmul(grad_avg, grad_avg, value: -1).sqrt!.add!(group[:eps])
             else
               avg = square_avg.sqrt.add!(group[:eps])
             end
 
             if group[:momentum] > 0
               buf = state[:momentum_buffer]
-              buf.mul!(group[:momentum]).addcdiv!(grad, avg)
-              p.data.add!(-group[:lr], buf)
+              buf.mul!(group[:momentum]).addcdiv!(grad, avg, value: 1)
+              p.data.add!(buf, alpha: -group[:lr])
             else
-              p.data.addcdiv!(-group[:lr], grad, avg)
+              p.data.addcdiv!(grad, avg, value: -group[:lr])
             end
           end
         end

data/lib/torch/optim/rprop.rb

@@ -52,7 +52,7 @@ module Torch
             grad[sign.eq(etaminus)] = 0
 
             # update parameters
-            p.data.addcmul!(-1, grad.sign, step_size)
+            p.data.addcmul!(grad.sign, step_size, value: -1)
 
             state[:prev].copy!(grad)
           end

data/lib/torch/optim/sgd.rb

@@ -36,14 +36,14 @@ module Torch
             end
             if momentum != 0
               param_state = @state[p]
-              if !param_state.key(:momentum_buffer)
+              if !param_state.key?(:momentum_buffer)
                 buf = param_state[:momentum_buffer] = Torch.clone(d_p).detach
               else
                 buf = param_state[:momentum_buffer]
                 buf.mul!(momentum).add!(d_p, alpha: 1 - dampening)
               end
               if nesterov
-                d_p = d_p.add(momentum, buf)
+                d_p = d_p.add(buf, alpha: momentum)
               else
                 d_p = buf
               end

data/lib/torch/tensor.rb

@@ -174,5 +174,10 @@ module Torch
       return _random!(0, *args) if args.size == 1
       _random!(*args)
     end
+
+    # center option
+    def stft(*args)
+      Torch.stft(*args)
+    end
   end
 end

data/lib/torch/version.rb

@@ -1,3 +1,3 @@
 module Torch
-  VERSION = "0.4.0"
+  VERSION = "0.5.2"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: torch-rb
 version: !ruby/object:Gem::Version
-  version: 0.4.0
+  version: 0.5.2
 platform: ruby
 authors:
 - Andrew Kane
-autorequire: 
+autorequire:
 bindir: bin
 cert_chain: []
-date: 2020-09-27 00:00:00.000000000 Z
+date: 2020-10-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rice
@@ -108,21 +108,7 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: 0.1.1
-- !ruby/object:Gem::Dependency
-  name: magro
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-description: 
+description:
 email: andrew@chartkick.com
 executables: []
 extensions:
@@ -288,7 +274,7 @@ homepage: https://github.com/ankane/torch.rb
 licenses:
 - BSD-3-Clause
 metadata: {}
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -303,8 +289,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.2
-signing_key: 
+rubygems_version: 3.1.4
+signing_key:
 specification_version: 4
 summary: Deep learning for Ruby, powered by LibTorch
 test_files: []