Merge pull request #9 from DaGaiBa/rc4main

Fix precision error of FocalLoss op

Author: momo609

mmcv/ops/csrc/pytorch/npu/focal_loss_npu.cpp

@@ -4,6 +4,13 @@ using namespace std;
 
 void sigmoid_focal_loss_forward_npu(Tensor input, Tensor target, Tensor weight,
                                     Tensor output, float gamma, float alpha) {
+  at::Tensor input_y = input;
+  at::Tensor output_y = output;
+  bool is_half = input.scalar_type() == at::kHalf;
+  if (is_half) {
+    input_y = input.to(at::kFloat);
+    output_y = output.to(at::kFloat);
+  }
   int64_t n_class = input.size(1);
   at::Tensor target_y = at::ones_like(input);
   if (n_class == 1) {
@@ -15,21 +22,28 @@ void sigmoid_focal_loss_forward_npu(Tensor input, Tensor target, Tensor weight,
   }
   target_y = target_y.to(at::kInt);
   int64_t weight_size = weight.size(0);
-  at::Tensor weight_y = at::ones_like(input);
+  at::Tensor weight_y = at::ones_like(input_y);
   if (weight_size > 0) {
     weight_y = at::broadcast_to(weight, input.sizes());
+    if (is_half) {
+      weight_y = weight_y.to(at::kFloat);
+    }
   }
   OpCommand cmd;
   string reduction = "none";
   cmd.Name("SigmoidFocalLoss")
-      .Input(input)
+      .Input(input_y)
       .Input(target_y)
       .Input(weight_y)
-      .Output(output)
+      .Output(output_y)
       .Attr("gamma", gamma)
       .Attr("alpha", alpha)
       .Attr("reduction", reduction)
       .Run();
+  if (is_half) {
+    output_y = output_y.to(at::kHalf);
+  }
+  output.copy_(output_y);
 }
 
 void sigmoid_focal_loss_forward_impl(Tensor input, Tensor target, Tensor weight,
@@ -38,6 +52,13 @@ void sigmoid_focal_loss_forward_impl(Tensor input, Tensor target, Tensor weight,
 void sigmoid_focal_loss_backward_npu(Tensor input, Tensor target, Tensor weight,
                                      Tensor grad_input, float gamma,
                                      float alpha) {
+  at::Tensor input_y = input;
+  at::Tensor grad_input_y = grad_input;
+  bool is_half = input.scalar_type() == at::kHalf;
+  if (is_half) {
+    input_y = input.to(at::kFloat);
+    grad_input_y = grad_input.to(at::kFloat);
+  }
   int64_t n_class = input.size(1);
   at::Tensor target_y = at::ones_like(input);
   if (n_class == 1) {
@@ -50,22 +71,29 @@ void sigmoid_focal_loss_backward_npu(Tensor input, Tensor target, Tensor weight,
   target_y = target_y.to(at::kInt);
   at::Tensor grad_up = at::ones_like(input);
   int64_t weight_size = weight.size(0);
-  at::Tensor weight_y = at::ones_like(input);
+  at::Tensor weight_y = at::ones_like(input_y);
   if (weight_size > 0) {
     weight_y = at::broadcast_to(weight, input.sizes());
+    if (is_half) {
+      weight_y = weight_y.to(at::kFloat);
+    }
   }
   OpCommand cmd;
   string reduction = "none";
   cmd.Name("SigmoidFocalLossGrad")
-      .Input(input)
+      .Input(input_y)
       .Input(target_y)
       .Input(grad_up)
       .Input(weight_y)
-      .Output(grad_input)
+      .Output(grad_input_y)
       .Attr("gamma", gamma)
       .Attr("alpha", alpha)
       .Attr("reduction", reduction)
       .Run();
+  if (is_half) {
+    grad_input_y = grad_input_y.to(at::kHalf);
+  }
+  grad_input.copy_(grad_input_y);
 }
 
 void sigmoid_focal_loss_backward_impl(Tensor input, Tensor target,
@@ -74,26 +102,38 @@ void sigmoid_focal_loss_backward_impl(Tensor input, Tensor target,
 
 void softmax_focal_loss_forward_npu(Tensor input, Tensor target, Tensor weight,
                                     Tensor output, float gamma, float alpha) {
+  at::Tensor input_y = input;
+  bool is_half = input.scalar_type() == at::kHalf;
+  if (is_half) {
+    input_y = input.to(at::kFloat);
+  }
   int64_t n_class = input.size(1);
   at::Tensor target_y = at::one_hot(target, n_class);
   target_y = target_y.to(at::kInt);
   int64_t weight_size = weight.size(0);
-  at::Tensor weight_y = at::ones_like(input);
+  at::Tensor weight_y = at::ones_like(input_y);
   if (weight_size > 0) {
     weight_y = at::broadcast_to(weight, input.sizes());
+    if (is_half) {
+      weight_y = weight_y.to(at::kFloat);
+    }
   }
-  at::Tensor op_output = at::ones_like(input);
+  
+  at::Tensor op_output = at::ones_like(input_y);
   OpCommand cmd;
   string reduction = "none";
   cmd.Name("SoftmaxFocalLoss")
-      .Input(input)
+      .Input(input_y)
       .Input(target_y)
       .Input(weight_y)
       .Output(op_output)
       .Attr("gamma", gamma)
       .Attr("alpha", alpha)
       .Attr("reduction", reduction)
       .Run();
+  if (is_half) {
+    op_output = op_output.to(at::kHalf);
+  }
   int64_t n_batch = input.size(0);
   c10::SmallVector<int64_t, 2> offsets = {0, 0};
   c10::SmallVector<int64_t, 2> sizes = {n_batch, 1};
@@ -124,27 +164,41 @@ void softmax_focal_loss_forward_impl(Tensor input, Tensor target, Tensor weight,
 void softmax_focal_loss_backward_npu(Tensor input, Tensor target, Tensor weight,
                                      Tensor buff, Tensor grad_input,
                                      float gamma, float alpha) {
+  at::Tensor input_y = input;
+  at::Tensor grad_input_y = grad_input;
+  bool is_half = input.scalar_type() == at::kHalf;
+  if (is_half) {
+    input_y = input.to(at::kFloat);
+    grad_input_y = grad_input.to(at::kFloat);
+  }
   int64_t n_class = input.size(1);
   at::Tensor target_y = at::one_hot(target, n_class);
   target_y = target_y.to(at::kInt);
   at::Tensor grad_up = at::ones_like(input);
   int64_t weight_size = weight.size(0);
-  at::Tensor weight_y = at::ones_like(input);
+  at::Tensor weight_y = at::ones_like(input_y);
   if (weight_size > 0) {
     weight_y = at::broadcast_to(weight, input.sizes());
+    if (is_half) {
+      weight_y = weight_y.to(at::kFloat);
+    }
   }
   OpCommand cmd;
   string reduction = "none";
   cmd.Name("SoftmaxFocalLossGrad")
-      .Input(input)
+      .Input(input_y)
       .Input(target_y)
       .Input(grad_up)
       .Input(weight_y)
-      .Output(grad_input)
+      .Output(grad_input_y)
       .Attr("gamma", gamma)
       .Attr("alpha", alpha)
       .Attr("reduction", reduction)
       .Run();
+  if (is_half) {
+    grad_input_y = grad_input_y.to(at::kHalf);
+  }
+  grad_input.copy_(grad_input_y);
 }
 
 void softmax_focal_loss_backward_impl(Tensor input, Tensor target,