Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Codegen for Tanh #3724

Merged
merged 1 commit into from
Jul 19, 2022
Merged

Codegen for Tanh #3724

merged 1 commit into from
Jul 19, 2022

Conversation

steventk-g
Copy link
Collaborator

Generated LazyIr.h:

class Tanh : public XlaNode {
 public:
  static torch::lazy::OpKind ClassOpKind() {
    return torch::lazy::OpKind(at::aten::tanh);
  }

  Tanh(const torch::lazy::Value& self, std::vector<torch::lazy::Shape>&& shapes)
      : XlaNode(torch::lazy::OpKind(at::aten::tanh),
              {self}, std::move(shapes),
              [&]() { return TanhOutputShape(self); },
              /* num_outputs */ 1,
              torch::lazy::MHash())
  {
    
  }

  std::string ToString() const override {
    std::stringstream ss;
    ss << XlaNode::ToString();
    
    return ss.str();
  }

  

  bool CanBeReused(const torch::lazy::Value& self) const {
    return false;
    }

  torch_xla::XlaOpVector Lower(LoweringContext* loctx) const override;

  
  

};

Generated XLANativeFunctions.cpp:

    at::Tensor XLANativeFunctions::tanh(const at::Tensor & self) {
        
        XLA_FN_COUNTER("xla::");
        auto common_device = torch_xla::bridge::GetXlaDevice(self);
        TORCH_INTERNAL_ASSERT(common_device);
        
        torch_xla::XLATensorPtr lazy_self = torch_xla::bridge::GetXlaTensorOrCreateForWrappedNumber(self, *common_device);
        torch::lazy::NodePtr node = torch::lazy::ReuseNode<Tanh>(lazy_self->GetIrValue());
        if (!node) {
                    auto self_meta = to_meta(self);
        auto out_meta = at::meta::tanh(self_meta);
        
std::vector<torch::lazy::Shape> shapes{torch::lazy::Shape(out_meta.scalar_type(), out_meta.sizes().vec())};
            TORCH_INTERNAL_ASSERT(shapes.size() == 1);
            if(torch::lazy::symbolicShapeEnabled()){
                std::vector<torch::jit::IValue> inputs = { self };
                const char* schema_str = "aten::tanh(Tensor self) -> Tensor";
                applySymbolicShapesOnLT(schema_str, inputs, shapes);
            }
        
            node = torch::lazy::MakeNode<Tanh>(lazy_self->GetIrValue(), std::move(shapes));
            CacheNode(node);
        }
        
        auto result = torch_xla::bridge::AtenFromXlaTensor(
                torch_xla::XLATensor::Create(std::move(node), *common_device));
        return result;
    };

@steventk-g steventk-g force-pushed the steventk-first-codegen branch from 1f4b697 to 68d8b0f Compare July 19, 2022 00:18
@steventk-g steventk-g force-pushed the steventk-first-codegen branch from 98d64ad to 966ff1a Compare July 19, 2022 00:35
Copy link
Collaborator

@wonjoolee95 wonjoolee95 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nice! Thanks! 👍

@wonjoolee95
Copy link
Collaborator

@steventk-g, also curious what was the cause of the problem you mentioned before regarding the Node being not found?

@JackCaoG JackCaoG merged commit 1262dd4 into master Jul 19, 2022
@JackCaoG JackCaoG deleted the steventk-first-codegen branch July 19, 2022 17:00
@steventk-g
Copy link
Collaborator Author

steventk-g commented Jul 19, 2022

@steventk-g, also curious what was the cause of the problem you mentioned before regarding the Node being not found?

I hadn't removed the Tanh definition from ops.h, so it was interfering with the template somehow. It wasn't clear from the compiler error, though. Once I removed it everything built fine.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants