RFC-0043 Torch Native Runtime #72

zhxchen17 · 2025-04-23T18:43:04Z

No description provided.

zhxchen17 · 2025-04-23T18:44:27Z

RFC-0043-torch-native-runtime.md

gmagogsfm · 2025-04-24T02:07:18Z

RFC-0043-torch-native-runtime.md

+
+### “Any performance regression from TorchScript?”
+
+* Single digit percentage (&lt;10%) improvement over TorchScript interpreter from improvements like static dispatch, memory planning, inter-op parallelism and overhead reduction.


Suggested change

* Single digit percentage (<10%) improvement over TorchScript interpreter from improvements like static dispatch, memory planning, inter-op parallelism and overhead reduction.

No, on the contrary we expect * single digit percentage (<10%) improvement over TorchScript interpreter from improvements like static dispatch, memory planning, inter-op parallelism and overhead reduction.

RFC-0043-torch-native-runtime.md

Summary: Torch Native Runtime RFC: pytorch/rfcs#72 Thi diff moves `TensorMeta.cpp` and `TensorMeta.h` to PyTorch core. Test Plan: ``` buck2 run mode/dev-nosan caffe2/test/cpp/nativert:nativert ``` Differential Revision: D73820548

Summary: Pull Request resolved: pytorch#152475 Torch Native Runtime RFC: pytorch/rfcs#72 Thi diff moves `TensorMeta.cpp` and `TensorMeta.h` to PyTorch core. Test Plan: ``` buck2 run mode/dev-nosan caffe2/test/cpp/nativert:nativert ``` Differential Revision: D73820548

Summary: Pull Request resolved: #152475 Torch Native Runtime RFC: pytorch/rfcs#72 Thi diff moves `TensorMeta.cpp` and `TensorMeta.h` to PyTorch core. Test Plan: Internal CI. GitHub CI with newly added test under `test/cpp/nativert/test_tensor_meta.cpp` Differential Revision: D73820548

Summary: Pull Request resolved: pytorch#152475 Torch Native Runtime RFC: pytorch/rfcs#72 Thi diff moves `TensorMeta.cpp` and `TensorMeta.h` to PyTorch core. Test Plan: Internal CI. GitHub CI with newly added test under `test/cpp/nativert/test_tensor_meta.cpp` Differential Revision: D73820548

Summary: Pull Request resolved: #152475 Torch Native Runtime RFC: pytorch/rfcs#72 Thi diff moves `TensorMeta.cpp` and `TensorMeta.h` to PyTorch core. Test Plan: Internal CI. GitHub CI with newly added test under `test/cpp/nativert/test_tensor_meta.cpp` Differential Revision: D73820548

Summary: Pull Request resolved: pytorch#152475 Torch Native Runtime RFC: pytorch/rfcs#72 Thi diff moves `TensorMeta.cpp` and `TensorMeta.h` to PyTorch core. Test Plan: Internal CI. GitHub CI with newly added test under `test/cpp/nativert/test_tensor_meta.cpp` Differential Revision: D73820548

Summary: Pull Request resolved: pytorch#152475 Torch Native Runtime RFC: pytorch/rfcs#72 Thi diff moves `TensorMeta.cpp` and `TensorMeta.h` to PyTorch core. Test Plan: Internal CI. GitHub CI with newly added test under `test/cpp/nativert/test_tensor_meta.cpp` Reviewed By: zhxchen17 Differential Revision: D73820548

Summary: Torch Native Runtime RFC: pytorch/rfcs#72 This diff moves `TensorMeta.cpp` and `TensorMeta.h` to PyTorch core under `torch/nativert/graph/` Existing `torch::_export::TensorMeta` in `torch/csrc/utils/generated_serialization_types.h` is auto-generated from the export serde schema and therefore only containing the most basic serializable types. We need the newly added `TensorMeta.cpp` to deserialize the metadata into a in-memory class with c10 types so that it can be consumed by the runtime later. Test Plan: Added test under `test/cpp/nativert/test_tensor_meta.cpp` Differential Revision: D73820548 Pull Request resolved: #152475 Approved by: https://github.com/albanD

Summary: Torch Native Runtime RFC: pytorch/rfcs#72 Port string utils functions join and split to c10/util Test Plan: Added tests in `string_util_test.cpp` buck2 run mode/opt caffe2/c10/test:util_base_tests Differential Revision: D74202473 Pull Request resolved: #152873 Approved by: https://github.com/cyyever, https://github.com/Skylion007

Summary: Torch Native Runtime RFC: pytorch/rfcs#72 Added an in-memory representation for input and output specs of a graph. The GraphSignature class models the input and output specs of an exported graph produced by torch.export, which holds the graph information deserialized from the pt2 archive package. Runtime relies on the GraphSignature for weight name lookup and weight loading. The serialization schema is defined in torch/_export/serde/schema.py See more at: https://docs.pytorch.org/docs/stable/export.html#torch.export.ExportGraphSignature Test Plan: Added tests under `test/cpp/nativert/test_graph_signature.cpp` Differential Revision: D73895378 Pull Request resolved: #152969 Approved by: https://github.com/swolchok

We split the large PR for adding Graph.h and Graph.cpp to `nativert` into 3 smaller PRs: 1. **Add header file** 2. Add source file 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75491860](https://our.internmc.facebook.com/intern/diff/D75491860/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. **Add source file** 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75492405](https://our.internmc.facebook.com/intern/diff/D75492405/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to `nativert` into 3 smaller PRs: 1. **Add header file** 2. Add source file 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75491860](https://our.internmc.facebook.com/intern/diff/D75491860/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. **Add source file** 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75492405](https://our.internmc.facebook.com/intern/diff/D75492405/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/) ghstack-source-id: 286574727 Pull Request resolved: #154532

… nativert" We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. **Add source file** 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75492405](https://our.internmc.facebook.com/intern/diff/D75492405/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to `nativert` into 3 smaller PRs: 1. **Add header file** 2. Add source file 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75491860](https://our.internmc.facebook.com/intern/diff/D75491860/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. **Add source file** 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75492405](https://our.internmc.facebook.com/intern/diff/D75492405/) [ghstack-poisoned]

Pull Request resolved: #154532 We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. ghstack-source-id: 286711904 @exported-using-ghexport Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/)

We split the large PR for adding Graph.h and Graph.cpp to `nativert` into 3 smaller PRs: 1. **Add header file** 2. Add source file 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75491860](https://our.internmc.facebook.com/intern/diff/D75491860/) [ghstack-poisoned]

Pull Request resolved: #154532 We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. ghstack-source-id: 286813445 @exported-using-ghexport Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/)

… nativert" We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. **Add source file** 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75492405](https://our.internmc.facebook.com/intern/diff/D75492405/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to `nativert` into 3 smaller PRs: 1. **Add header file** 2. Add source file 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75491860](https://our.internmc.facebook.com/intern/diff/D75491860/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. **Add source file** 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75492405](https://our.internmc.facebook.com/intern/diff/D75492405/) [ghstack-poisoned]

Pull Request resolved: #154532 #buildall We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. ghstack-source-id: 287076545 @exported-using-ghexport Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/)

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. **Add source file** 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75492405](https://our.internmc.facebook.com/intern/diff/D75492405/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to `nativert` into 3 smaller PRs: 1. **Add header file** 2. Add source file 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75491860](https://our.internmc.facebook.com/intern/diff/D75491860/) [ghstack-poisoned]

… nativert" We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. Add source file 3. **Add test and build rules** Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75495273](https://our.internmc.facebook.com/intern/diff/D75495273/) [ghstack-poisoned]

We split the large PR for adding Graph.h and Graph.cpp to nativert into 3 smaller PRs: 1. Add header file 2. **Add source file** 3. Add test and build rules. Torch Native Runtime RFC: pytorch/rfcs#72 4 classes have been introduced: `Graph`, `Node`, `Value`, `Type` - `Type` represents the kind of a `Value` - `Value` represents a single symbolic value, it could be any kind that exists in `Type`. Values are inputs and outputs of a `Node`. - `Node` represents a single unit of execution, typically a PyTorch op. - `Graph` represents a model's computation graph, which is designed to facilitate transformation/analysis. Differential Revision: [D75492405](https://our.internmc.facebook.com/intern/diff/D75492405/) [ghstack-poisoned]

RFC-0043 Torch Native Runtime

1175e55

facebook-github-bot added the cla signed label Apr 23, 2025

zhxchen17 mentioned this pull request Apr 23, 2025

RFC: Torch Native Runtime pytorch/pytorch#152034

Open

2 tasks

gmagogsfm reviewed Apr 24, 2025

View reviewed changes

zhxchen17 requested a review from gmagogsfm April 28, 2025 16:17

yiming0416 mentioned this pull request Apr 29, 2025

[nativert] Move TensorMeta to pytorch core pytorch/pytorch#152475

Closed

yiming0416 mentioned this pull request May 5, 2025

[nativert] Port string join and split to c10/util pytorch/pytorch#152873

Closed

yiming0416 mentioned this pull request May 6, 2025

[nativert] Move GraphSignature to pytorch core pytorch/pytorch#152969

Closed

Update RFC-0043-torch-native-runtime.md

fc0c9c2

yiming0416 mentioned this pull request May 21, 2025

[nativert] Move Graph to pytorch core pytorch/pytorch#153994

Open

yiming0416 mentioned this pull request May 28, 2025

[1/3] Add header file for Graph in nativert pytorch/pytorch#154528

Closed

yiming0416 mentioned this pull request May 28, 2025

[2/3] Add source file for Graph in nativert pytorch/pytorch#154529

Closed

yiming0416 mentioned this pull request May 28, 2025

[1/3] Add header file for Graph in nativert pytorch/pytorch#154530

Open

yiming0416 mentioned this pull request May 28, 2025

[2/3] Add source file for Graph in nativert pytorch/pytorch#154531

Open

yiming0416 mentioned this pull request May 28, 2025

[3/3] Add build rule and test for Graph in nativert pytorch/pytorch#154532

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

RFC-0043 Torch Native Runtime #72

RFC-0043 Torch Native Runtime #72

Uh oh!

zhxchen17 commented Apr 23, 2025

Uh oh!

zhxchen17 commented Apr 23, 2025

Uh oh!

Uh oh!

gmagogsfm Apr 24, 2025

Uh oh!

Uh oh!

Uh oh!


		### “Any performance regression from TorchScript?”

		* Single digit percentage (<10%) improvement over TorchScript interpreter from improvements like static dispatch, memory planning, inter-op parallelism and overhead reduction.

	* Single digit percentage (<10%) improvement over TorchScript interpreter from improvements like static dispatch, memory planning, inter-op parallelism and overhead reduction.
	No, on the contrary we expect * single digit percentage (<10%) improvement over TorchScript interpreter from improvements like static dispatch, memory planning, inter-op parallelism and overhead reduction.

RFC-0043 Torch Native Runtime #72

Are you sure you want to change the base?

RFC-0043 Torch Native Runtime #72

Uh oh!

Conversation

zhxchen17 commented Apr 23, 2025

Uh oh!

zhxchen17 commented Apr 23, 2025

Uh oh!

Uh oh!

gmagogsfm Apr 24, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!