Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / pytorch/TensorRT issues and pull requests

#2042 - ✨[Feature]Is there any way to prevent certain layers from being quantized to int8?

Issue - State: closed - Opened by sanbuphy over 1 year ago - 1 comment
Labels: feature request, No Activity

#2041 - 🐛 [Bug] Tensors on CPU/CUDA issue encountered in TorchScript

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment
Labels: bug, No Activity, component: partitioning

#2039 - FX converter documentation

Pull Request - State: closed - Opened by apbose over 1 year ago
Labels: documentation, cla signed

#2037 - Refactor dynamo and change default ir to dynamo

Issue - State: closed - Opened by peri044 over 1 year ago - 1 comment
Labels: No Activity

#2036 - 🐛 [Bug] Issue with `remove_ops` lowering pass in FX/Dynamo

Issue - State: closed - Opened by gs-olive over 1 year ago - 3 comments
Labels: bug, No Activity, component: fx, component: dynamo

#2035 - Setup continuous deployments on GHA

Issue - State: closed - Opened by narendasan over 1 year ago
Labels: infrastructure, Story: Porting CI/CD to GHA, Story: Infrastructure Upgrades

#2034 - 📖 [Story] Infrastructure Upgrades

Issue - State: closed - Opened by narendasan over 1 year ago - 1 comment
Labels: No Activity, infrastructure, Story

#2033 - Update Dynamo/TRT Workflow RFC

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment
Labels: component: dynamo

#2031 - ✨[Feature] Automatic type promotion for elementwise ops in FX

Issue - State: closed - Opened by gs-olive over 1 year ago - 2 comments
Labels: feature request, No Activity, component: dynamo

#2028 - ❓ [Question] Torch-TensorRT 1.3.0 uses cuDNN 8.6.0 instead of 8.5.0

Issue - State: open - Opened by akaimody123 over 1 year ago - 2 comments
Labels: question, No Activity

#2027 - 🐛 [Bug] Encountered bug when compiling facebook/segment_anything

Issue - State: open - Opened by fumin over 1 year ago - 1 comment
Labels: bug, component: converters, No Activity

#2026 - 🐛 [Bug] Using an ellipsis when indexing raises a RuntimeError

Issue - State: closed - Opened by airalcorn2 over 1 year ago - 2 comments
Labels: bug, component: converters, No Activity

#2025 - 🐛 [Bug] Torch-TensoRT does not write values to a Tensor when using non-consecutive indices

Issue - State: closed - Opened by airalcorn2 over 1 year ago - 2 comments
Labels: bug, component: converters, No Activity

#2024 - 🐛 [Bug] Torch-TensorRT silently converts a boolean mask into an integer index containing zeros and ones

Issue - State: closed - Opened by airalcorn2 over 1 year ago - 3 comments
Labels: bug, component: converters, No Activity

#2018 - 🐛 [Bug] Segmentation fault when using int8 precision

Issue - State: closed - Opened by AhmetHamzaEmra over 1 year ago - 4 comments
Labels: bug, No Activity, component: quantization

#2016 - 🐛 [Bug] Torch-TensorRT takes long time to compile densenet161

Issue - State: closed - Opened by Anlarry over 1 year ago - 3 comments
Labels: bug

#2009 - chore: remove compose_bmm pass as it is not decomposed anymore

Pull Request - State: closed - Opened by peri044 over 1 year ago - 1 comment
Labels: component: api [Python], cla signed, component: fx, fx

#2008 - ✨[Feature] Automatic casting of Long/Double Tensor constants in FX

Issue - State: open - Opened by gs-olive over 1 year ago - 1 comment
Labels: feature request, No Activity, fx, component: dynamo

#2007 - Unable to remove the QDQ after quantized Conv

Issue - State: closed - Opened by tp-nan over 1 year ago - 1 comment
Labels: No Activity, component: quantization

#1996 - ✨[Feature] Centralize Transposed Convolution Operations in FX

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment
Labels: feature request, component: fx, component: dynamo

#1995 - 🐛 [Bug] Encountered bug when compiling `fasterrcnn_mobilenet_v3_large_320_fpn` in Dynamo

Issue - State: closed - Opened by gs-olive over 1 year ago - 2 comments
Labels: bug, No Activity, component: dynamo

#1993 - ✨[Feature] Upgrade `setup.py` Python-only install

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment
Labels: feature request, component: build system, No Activity, component: dynamo

#1992 - 📖 [Story] TorchScript / Dynamo Feature Parity

Issue - State: closed - Opened by narendasan over 1 year ago - 1 comment
Labels: No Activity, Story

#1991 - fix 2 level nested conditional missing value bug

Pull Request - State: closed - Opened by bowang007 over 1 year ago - 1 comment
Labels: component: core, component: partitioning, cla signed

#1987 - 🐛 [Bug] Support dynamic shape for aten::chunk converter

Issue - State: closed - Opened by bowang007 over 1 year ago - 1 comment
Labels: bug, No Activity

#1986 - BigVGAN vocoder model working with trtexec but not via torch-trt

Issue - State: open - Opened by Christina-Young-NVIDIA over 1 year ago - 3 comments
Labels: No Activity

#1986 - BigVGAN vocoder model working with trtexec but not via torch-trt

Issue - State: closed - Opened by Christina-Young-NVIDIA over 1 year ago - 3 comments
Labels: No Activity

#1985 - Centralize converters for `MaxPool1D` and `einsum`

Issue - State: closed - Opened by gs-olive over 1 year ago - 5 comments
Labels: component: fx, component: dynamo

#1984 - Add example scripts for module/function level acceleration

Issue - State: closed - Opened by gs-olive over 1 year ago - 6 comments
Labels: No Activity, Story: Export/Compile Unification, Story: Dynamo Compile Improvements

#1974 - Upstream main to release 1.4

Pull Request - State: closed - Opened by narendasan over 1 year ago
Labels: component: tests, component: lowering, component: conversion, component: core, component: converters, component: api [Python], component: api [C++], component: evaluators, cla signed, component: fx, fx

#1973 - fix: Add new TRT 8.6 features to Dynamo compile [2 / x]

Pull Request - State: open - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: dynamo, Story: Export/Compile Unification, Story: Dynamo Compile Improvements

#1972 - fix: Allow FX convolution layers to take bias inputs

Pull Request - State: open - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, component: dynamo, Story: Dynamo Compile Improvements

#1971 - fix: Add new TRT 8.6 features to Dynamo compile

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: dynamo, Story: Export/Compile Unification, Story: Dynamo Compile Improvements

#1970 - fix: Rewrite constant_pad_nd to use a single slice layer for performance

Pull Request - State: closed - Opened by mfeliz-cruise over 1 year ago
Labels: component: conversion, component: core, component: converters, cla signed

#1969 - fix: Add support for `truncate_long_and_double` in Dynamo

Pull Request - State: open - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: dynamo, Story: Dynamo Compile Improvements

#1968 - feat/fix: Add support for general-purpose function exclusion in Dynamo

Pull Request - State: open - Opened by gs-olive over 1 year ago
Labels: WIP, component: api [Python], cla signed, component: dynamo, Story: Dynamo Compile Improvements

#1967 - Reorg docs and fix css for sphix-gallery examples

Pull Request - State: open - Opened by narendasan over 1 year ago
Labels: documentation, cla signed

#1966 - examples: Add example usage scripts for `torch_tensorrt.dynamo.compile` path [1 / x]

Pull Request - State: open - Opened by gs-olive over 1 year ago
Labels: documentation, component: build system, component: api [Python], cla signed, component: dynamo, Story: Dynamo Compile Improvements

#1965 - ATen converters refactor: squeeze/unsqueeze, permute, transpose, clamp

Pull Request - State: open - Opened by borisfom over 1 year ago - 2 comments
Labels: component: api [Python], cla signed, component: fx, fx

#1964 - ✨[Feature] Add support for `truncate_long_and_double` in Dynamo compile

Issue - State: open - Opened by gs-olive over 1 year ago
Labels: feature request, Story: Dynamo Compile Improvements

#1963 - ✨[Feature]Lowering pass for reciprocal

Issue - State: closed - Opened by apbose over 1 year ago
Labels: feature request

#1962 - ✨[Feature]Implement sqrt

Issue - State: closed - Opened by apbose over 1 year ago - 1 comment
Labels: feature request

#1961 - Reciprocal dynamo pass

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx

#1960 - [19/N] Moving unary core to impl - sqrt(FX Converter Refactor [8/N])<Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx

#1959 - ✨[Feature] Exclude key functions from tracing during AOT

Issue - State: closed - Opened by gs-olive over 1 year ago - 2 comments
Labels: feature request, No Activity, Story: Dynamo Compile Improvements

#1958 - WIP : Refactoring acc_op converters, adding new aten converters

Pull Request - State: closed - Opened by borisfom over 1 year ago - 3 comments
Labels: component: api [Python], cla signed, component: fx, fx

#1957 - fix: int/int=float division

Pull Request - State: closed - Opened by mfeliz-cruise over 1 year ago
Labels: component: tests, component: conversion, component: core, component: converters, cla signed

#1956 - cherry-pick: Dynamo upgrades and bugfixes (`release/1.4`)

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, release: v1.4

#1955 - fix: Add support for fake tensors

Pull Request - State: open - Opened by gs-olive over 1 year ago
Labels: WIP, component: api [Python], cla signed, component: fx, fx, component: dynamo, Story: Dynamo Compile Improvements

#1954 - ✨[Feature] + 🐛 [Bug] Allow `ITensor` biases in `aten.convolution` converters

Issue - State: open - Opened by gs-olive over 1 year ago
Labels: feature request, Story: ATen Op Support

#1953 - fix: Add decomposition for `aten.addmm`

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: dynamo

#1952 - feat: Add support for passing through build issues in Dynamo compile

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed

#1951 - ✨[Feature] Fully support `FakeTensor`s in Dynamo compile

Issue - State: open - Opened by gs-olive over 1 year ago
Labels: feature request, Story: Dynamo Compile Improvements

#1950 - Relaxing glob pattern for CUDA12

Pull Request - State: closed - Opened by borisfom over 1 year ago
Labels: cla signed

#1949 - fix: Address `.numpy()` issue on fake tensors

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: dynamo

#1947 - fix: Support dims < -1 in aten::stack converter

Pull Request - State: closed - Opened by mfeliz-cruise over 1 year ago
Labels: component: tests, component: conversion, component: core, component: converters, cla signed

#1946 - 📖 [Story] Porting Testing to GHA

Issue - State: open - Opened by narendasan over 1 year ago - 1 comment
Labels: No Activity, infrastructure, Story, Story: Porting CI/CD to GHA, Story: Infrastructure Upgrades

#1945 - fix: Add lowering pass to remove output repacking in `convert_method_to_trt_engine` calls

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: component: tests, component: lowering, component: core, component: api [Python], component: api [C++], cla signed

#1943 - ✨[Feature] Remove C++ Dependency in Dynamo

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment
Labels: feature request, Story: Export/Compile Unification, Story: Dynamo Compile Improvements

#1941 - 📖 [Story] Improve the `torch_tensorrt.dynamo.compile` path

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment
Labels: Story, Story: Dynamo Compile Improvements

#1940 - 📖 [Story] Sharing Components between Torch `export` and `compile` Paths in Torch-TRT

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment
Labels: Story: Export/Compile Unification

#1938 - ✨[Feature] Automatically remove output repacking for `convert_method_to_trt_engine` calls

Issue - State: closed - Opened by gs-olive over 1 year ago
Labels: feature request

#1937 - fix/feat: Add lowering pass to resolve most `aten::Int.Tensor` uses

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: component: tests, component: lowering, component: core, cla signed

#1935 - Upstream 3 features to fx_ts_compat: MS, VC, Optimization Level

Pull Request - State: closed - Opened by wu6u3tw over 1 year ago - 5 comments
Labels: component: api [Python], cla signed

#1933 - Fix: fix the bug that uninitialized tensor cannot be found

Pull Request - State: closed - Opened by bowang007 over 1 year ago - 2 comments
Labels: component: tests, component: core, component: partitioning, cla signed

#1932 - 🐛 [Bug] Segfault when trying to export detectron2 model (GeneralizedRCNN) to Torch-TensorRT

Issue - State: closed - Opened by anandrajaraman over 1 year ago - 2 comments
Labels: bug, No Activity

#1929 - ✨[Feature] Add Support for TRT 8.6 Hardware Compatibility Mode

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment
Labels: feature request

#1927 - ✨[Feature] Allow "pass-through" of engine-build errors in Dynamo

Issue - State: closed - Opened by gs-olive over 1 year ago
Labels: feature request, Story: Dynamo Compile Improvements

#1926 - Upstream 3 features for torch-trt: ms, VC, and optimisation level

Pull Request - State: closed - Opened by wu6u3tw over 1 year ago - 9 comments
Labels: component: api [Python], cla signed, component: fx, fx

#1925 - AttributeError: 'WeightNorm' object has no attribute '__name__'

Issue - State: open - Opened by gs-olive over 1 year ago - 1 comment
Labels: No Activity

#1924 - Refactor FX Documentation for Website

Issue - State: open - Opened by gs-olive over 1 year ago - 1 comment

#1924 - Refactor FX Documentation for Website

Issue - State: closed - Opened by gs-olive over 1 year ago - 1 comment

#1923 - Testing Time Increase in FX `acc` tests for Python TRT 8.6.1

Issue - State: open - Opened by gs-olive over 1 year ago
Labels: bug, bug: triaged [verified], release: v1.4

#1923 - Testing Time Increase in FX `acc` tests for Python TRT 8.6.1

Issue - State: open - Opened by gs-olive over 1 year ago
Labels: bug, bug: triaged [verified], release: v1.4

#1922 - Moving normalization core to impl - where(FX Converter Refactor [18/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1922 - Moving normalization core to impl - where(FX Converter Refactor [18/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1921 - feat: Prototype Module-Acceleration in Dynamo

Pull Request - State: open - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: dynamo, Story: Dynamo Compile Improvements

#1920 - how to convert itensor to pytorch tensor in torch-tensorrt fx mode?

Issue - State: closed - Opened by shuyuan-wang over 1 year ago - 4 comments
Labels: No Activity

#1919 - 🐛 [Bug] Encountered bug when using Torch-TensorRT to run a PyTorch model

Issue - State: closed - Opened by terU3760 over 1 year ago - 3 comments
Labels: bug, No Activity

#1918 - fix: Add torchvision legacy CI parameter

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: cla signed

#1917 - Moving elementwise core to impl - matmul(FX Converter Refactor [17/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1917 - Moving elementwise core to impl - matmul(FX Converter Refactor [17/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1916 - Moving slice to impl - slice (FX Converter Refactor [16/N]) - <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1916 - Moving slice to impl - slice (FX Converter Refactor [16/N]) - <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1915 - Moving select to impl - select(FX Converter Refactor [15/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago - 1 comment
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1914 - Moving squeeze to impl - squeeze (FX Converter Refactor [14/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1913 - Reorg for converters gelu (FX Converter Refactor [13/N])

Pull Request - State: open - Opened by apbose over 1 year ago - 1 comment
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1913 - Reorg for converters gelu (FX Converter Refactor [13/N]) <Target: converter_reorg_proto>

Pull Request - State: open - Opened by apbose over 1 year ago - 1 comment
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1912 - ❓ [Question] How to correctly convert model by using torch-tensorrt

Issue - State: open - Opened by HtutLynn over 1 year ago - 1 comment
Labels: question, No Activity

#1912 - ❓ [Question] How to correctly convert model by using torch-tensorrt

Issue - State: open - Opened by HtutLynn over 1 year ago
Labels: question

#1911 - feat: support type promotion in aten::cat converter

Pull Request - State: open - Opened by mfeliz-cruise over 1 year ago
Labels: component: tests, component: conversion, component: core, component: converters, cla signed

#1910 - fix: Cherry-pick `acc` convolution fix to `release/1.4`

Pull Request - State: closed - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, release: v1.4

#1910 - fix: Cherry-pick `acc` convolution fix to `release/1.4`

Pull Request - State: open - Opened by gs-olive over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, release: v1.4

#1909 - Moving normalization core to impl - softmax (FX Converter Refactor [12/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1908 - Moving elementwise nornalization to impl- layer_norm (FX Converter Refactor [11/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1907 - Moving normalization core to impl - batch_norm(FX Converter Refactor [10/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support

#1906 - Moving elementwise core to impl - rsub(FX Converter Refactor [4/N]) <Target: converter_reorg_elementwise>

Pull Request - State: open - Opened by apbose over 1 year ago
Labels: component: api [Python], cla signed, component: fx, fx, Story: ATen Op Support