id int64 393k 2.82B | repo stringclasses 68
values | title stringlengths 1 936 | body stringlengths 0 256k ⌀ | labels stringlengths 2 508 | priority stringclasses 3
values | severity stringclasses 3
values |
|---|---|---|---|---|---|---|
885,216,343 | flutter | [flutter_markdown] MarkdownBody widgets generate links which violate kMinInteractiveDimension | I'm using flutter_markdown 0.6.1 and Flutter version 2.0.6, and I really appreciate the package! It's honestly a life saver when rendering user-generated content.
When I use MarkdownBody to render markdown text with links, I notice that the links are often not large enough to successfully tap on the first try. Depen... | package,team-ecosystem,has reproducible steps,P3,found in release: 2.0,found in release: 2.3,p: flutter_markdown,triaged-ecosystem | low | Critical |
885,247,262 | rust | Bad error messages when passing and using 2 PartialEq or PartialOrd traits to function. | Given the following code:
```rust
<#![allow(unused)]
use std::cmp::Ordering;
fn main() {
let m1 = MyStruct(0);
let m2 = MyStruct(1);
println!("{:?}",m1 == m2);
println!("{:?}",m1 > m2);
}
fn test_eq(m1 :&impl PartialEq, m2 :&impl PartialEq) {
println!("{:?}",m1 == m2);
}... | A-diagnostics,T-compiler | low | Critical |
885,298,636 | rust | Automatic &mut reborrowing doesn't happen when arguments are for `impl Trait` | Given code like this:
```rust
fn foo(a: &mut i32) {
auto_reborrow(a); // works
auto_reborrow(a); // works
no_reborrow(a); // works
no_reborrow(a); // ERROR
}
fn auto_reborrow(_: &mut i32) {}
fn no_reborrow(_: impl std::fmt::Debug) {}
```
it [currently fails to compile](https://play.... | A-diagnostics,T-lang,A-inference,D-confusing | low | Critical |
885,453,383 | go | runtime/race: panic+recover in init() adds extra info to data race stack traces in main() | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.16.4 windows/amd64
</pre>
### Does this issue reprodu... | RaceDetector,NeedsInvestigation,compiler/runtime | low | Critical |
885,460,998 | pytorch | torch.distributed.nn.all_reduce incorrectly scales the gradient | ## 🐛 Bug
`torch.distributed.nn.all_reduce` computes different gradient values from `torch.distributed.all_reduce`. In particular, it seems to scale the gradients by `world_size` incorrectly.
## To Reproduce
Script to reproduce the behavior:
```python
import torch
import torch.distributed
import torch.di... | high priority,oncall: distributed,triaged | medium | Critical |
885,565,586 | rust | Following compiler suggestions for fixing errors sends me in a loop between tuple syntax and struct syntax. | I tried matching on the wrong error type. The compiler suggested I use a struct syntax instead of the tuple syntax. When I tried the suggested struct syntax it told me to use the same tuple syntax it had just told me not to use. I've been able to boil it down to a simpler repro case below. This is an analog of the fixe... | A-diagnostics,T-compiler,C-bug,A-suggestion-diagnostics | low | Critical |
886,016,899 | pytorch | Support a dist.group_like API | ## 🚀 Feature
Sometimes it is useful to create a process group similar to another process group, but with a different backend. For example, this came up when using DDP for dense parameter synchronization and having to create a gloo pg with the same set of ranks for DDP across the sparse part (since NCCL doesn't yet ... | oncall: distributed,module: bootcamp,triaged,pt_distributed_rampup,module: c10d | low | Minor |
886,074,079 | pytorch | Eliminate potential double device check | #56872 adds device check support in dispatcher wrapper. See https://dev-discuss.pytorch.org/t/device-check-will-be-automatically-generated-for-functions/211 for details.
For kernels implement its own device check, this results in checking tensor device twice. We should:
1. Remove device check in kernel if possibl... | triaged,better-engineering | low | Minor |
886,304,395 | material-ui | [icons] Not able to install icons yarn add @material-ui/icons package | ## Current Behavior 😯
<!-- Describe what happens instead of the expected behavior. -->
> info There appears to be trouble with your network connection. Retrying...
## Expected Behavior 🤔
<!-- Describe what should happen. -->
## Steps to Reproduce 🕹
<!--
Provide a link to a live example (you can... | new feature,performance,package: icons | low | Critical |
886,357,593 | rust | Unhelpful error on implicit `'static` lifetime for `dyn Trait` | Given [the following code](https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=b43430a225505f0ff53d18cfc06dae25):
```rust
struct S(Vec<Box<dyn std::fmt::Debug>>);
impl S {
fn update(&mut self) {
let iter = self.0.iter_mut().map(|p| &mut **p);
use_it(iter);
}
}
fn u... | A-diagnostics,T-compiler | low | Critical |
886,377,095 | rust | Lifetime coercion with lifetime bounds on traits | I had lifetime errors with `async_trait`. I made a minimal example which triggers the same error message:
```rust
fn main() {}
trait Worker {
fn work<'a, 'b>(input: &'a mut String) -> &'b String
where
'a: 'b;
}
trait WorkerCoerced {
fn work<'a>(input: &'a mut String) -> &'a String;
}... | T-lang,C-bug | low | Critical |
886,553,191 | flutter | Translucent Flutter Screen glitches on Application Resume | https://user-images.githubusercontent.com/10841769/117787685-44ccee00-b279-11eb-91db-878c4ddc1856.mp4
When the app goes back to the background and back again, the transparency of the background color is lost！
Here is jump code with translucent！
<details>
<summary>code snippet</summary>
```dart
currentAc... | platform-android,engine,has reproducible steps,P2,found in release: 2.0,found in release: 2.3,team-android,triaged-android | low | Minor |
886,626,452 | material-ui | [Stepper] Long labels break step connectors | <!-- Provide a general summary of the issue in the Title above -->
<!--
Thank you very much for contributing to Material-UI by creating an issue! ❤️
To avoid duplicate issues we ask you to check off the following list.
-->
<!-- Checked checkbox should look like this: [x] -->
- [x] The issue is present i... | waiting for 👍,component: stepper | low | Critical |
886,699,932 | go | x/text/cases: incorrect case folding for Ꮭ | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.16 darwin/amd64
</pre>
### Does this issue reproduce ... | NeedsInvestigation | low | Critical |
886,833,146 | pytorch | [feature request] torch.as_tensor to support any object that NumPy's asarray or array can consume (consume __array_interface__) | This would enlarge the accepted set of inputs of `torch.as_tensor` and would support PIL images / h5py arrays. I think this feature request goes well in the theme of standardizing support for methods like `__array_interface__`, `__cuda_array_interface__` and such. It would be good for `torch.as_tensor` to support accep... | triaged,module: numpy | low | Critical |
886,904,472 | pytorch | linalg.eigh and linalg.cholesky UPLO flag breaks in backward | ## 🐛 (Explanation of the) Bug
tl;dr The UPLO flag inherited fro Numpy is an implementation detail that is exposed to allow for some memory optimisations when dealing with symmetric matrices. This optimisations lead to incorrect gradients.
The UPLO flag in `torch.linalg.eigh` and `torch.linalg.cholesky` has been in... | module: autograd,triaged,module: linear algebra | low | Critical |
886,924,845 | flutter | Add components version on LicenseRegistry | <!-- Thank you for using Flutter!
If you are looking for support, please check out our documentation
or consider asking a question on Stack Overflow:
* https://flutter.dev/
* https://api.flutter.dev/
* https://stackoverflow.com/questions/tagged/flutter?sort=frequent
If you hav... | c: new feature,framework,c: proposal,P3,team-framework,triaged-framework | low | Critical |
887,081,844 | TypeScript | Can references path support package name? | # Suggestion
<!--
Please fill in each section completely. Thank you!
-->
## 🔍 Search Terms
<!--
💡 Did you know? TypeScript has over 2,000 open suggestions!
🔎 Please search thoroughly before logging new feature requests as most common ideas already have a proposal in progress.
The "Common Featur... | Suggestion,Awaiting More Feedback | low | Major |
887,150,028 | excalidraw | Freedraw exported svg can contain NaN | https://excalidraw.com/#json=5110389188591616,PfDJnUAPfTuaMyALYyEY6Q
```xml
<svg version="1.1" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 32.078125 86.953125" width="64.15625" height="173.90625">
<!-- svg-source:excalidraw -->
<defs>
<style>
@font-face {
font-family: "Virgil";
... | bug,freedraw | low | Minor |
887,250,625 | material-ui | outlined TextField with long label causes scroll | <!-- Provide a general summary of the issue in the Title above -->
<!--
Thank you very much for contributing to Material-UI by creating an issue! ❤️
To avoid duplicate issues we ask you to check off the following list.
-->
<!-- Checked checkbox should look like this: [x] -->
- [x] The issue is present i... | component: text field | low | Critical |
887,277,307 | PowerToys | Add Snap Assist Windows preview tool to FZ | Add snap assist to fancy zones! | Idea-Enhancement,Product-FancyZones | low | Minor |
887,352,648 | flutter | Hide keyboard from system in input and editabletext | When using any type of input, from a TexField to an EditableText, you cannot choose that the system keyboard does not appear, whether you focus on the input or select the text. I am trying to make a calculator and I cannot make my text editable because I cannot hide the system keyboard.
I propose to add the option t... | a: text input,c: new feature,framework,a: typography,c: proposal,P3,team-framework,triaged-framework | low | Minor |
887,635,021 | material-ui | [ESM][core] Add `exports` to the package.json | <!-- Provide a general summary of the feature in the Title above -->
<!--
Thank you very much for contributing to Material-UI by creating an issue! ❤️
To avoid duplicate issues we ask you to check off the following list.
-->
<!-- Checked checkbox should look like this: [x] -->
- [x] I have searched the ... | breaking change,enhancement,scope: code-infra | medium | Major |
887,668,565 | go | x/build/cmd/coordinator: failed to add requested SlowBot | In https://go-review.googlesource.com/c/go/+/318570/1#message-c05120c76c77004b072d84595dbb73a0392f90a4, I added a `TRY=longtest` comment and `Run-TryBot+1` to a CL.
I had previously removed a `Run-TryBot+1` and a `TryBot-Result` for the same patch set after the TryBot run had flaked.
Unfortunately, the coordinato... | Builders,NeedsInvestigation | low | Critical |
887,744,287 | pytorch | quantize_per_tensor not symbolically traceable with FX if scale+zp are proxied | ## 🐛 Bug
Unable to symbolically trace a `torch.quantize_per_tensor` method call with FX if scale and zero_point are proxied.
## To Reproduce
```
# code
import torch
from torch.quantization.quantize_fx import prepare_fx, convert_fx
from torch.fx.expe... | triaged,module: fx | low | Critical |
887,769,959 | pytorch | PyTorch master failed to build with glog 0.5.0. | ## Bug
Log:
```
root@164a9000a4ab:/tmp/scratch/pytorch/build# cmake --build . -- -j1
[1/153] : && /usr/bin/g++-8 -fdebug-prefix-map='/tmp/scratch'='/usr/local/src' -g -march=haswell -mtune=generic -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTOR... | module: build,triaged | medium | Critical |
887,829,615 | TypeScript | lib.dom.d.ts should escape html elements in jsdoc comments | # lib Update Request
<!--
Please fill in each section completely. Thank you!
Are you here for one of these commonly-requested lib changes?
* Object.keys - see https://stackoverflow.com/questions/55012174/
* Array methods - see https://github.com/microsoft/TypeScript/issues/36554
* parseInt, pa... | Suggestion,Help Wanted,Experience Enhancement | low | Major |
887,866,289 | transformers | Routing Transformers / Add Google PG-19 Models | # 🌟 New model addition - Google PG-19 Models
## Model description
Model checkpoints finally released as discussed in "Efficient Content-Based Sparse Attention with Routing Transformers'
Aurko Roy, Mohammad Saffar, Ashish Vaswani, David Grangier (https://arxiv.org/abs/2003.05997)
## Open source status
* [X... | New model | low | Minor |
888,152,752 | go | all: replace x.Type().Kind() to x.Kind() | If `x` is a `reflect.Value` value and it is not the zero `Value`, `x.Kind()` can be used instead of `x.Type().Kind()`.
There are many places in the standard library where this replacement can be done
```
% grep -r ".Type().Kind()" src
src/cmd/fix/cftype.go: if v.Type().Kind() != reflect.Ptr {
src/cmd/fix/cft... | Performance,NeedsInvestigation | low | Critical |
888,241,627 | vscode | Render right or left bracket border thicker depending on the cursor position relative to initial bracket | <!-- ⚠️⚠️ Do Not Delete This! feature_request_template ⚠️⚠️ -->
<!-- Please read our Rules of Conduct: https://opensource.microsoft.com/codeofconduct/ -->
<!-- Please search existing issues to avoid creating duplicates. -->
<!-- Describe the feature you'd like. -->
# Discussion
Sometimes when you are writing... | feature-request,editor-bracket-matching | low | Major |
888,441,544 | vscode | Applying multiple valid workspace edits and then awaiting after can fail unexpectedly. | <!-- ⚠️⚠️ Do Not Delete This! bug_report_template ⚠️⚠️ -->
<!-- Please read our Rules of Conduct: https://opensource.microsoft.com/codeofconduct/ -->
<!-- 🕮 Read our guide about submitting issues: https://github.com/microsoft/vscode/wiki/Submitting-Bugs-and-Suggestions -->
<!-- 🔎 Search existing issues to avoid cr... | under-discussion,workspace-edit | low | Critical |
888,448,902 | TypeScript | TS intellisense getting tripped on const object indexes | *TS Template added by @mjbvz*
**TypeScript Version**: 4.4.0-dev.20210511
**Search Terms**
- suggest / suggestions
- Bracket accessor
---
Issue Type: <b>Bug</b>
```ts
const a = {
foo: 'foo',
} as const;
const c = {
[a.foo]: 'foo',
} as const;
c.
```
Suggests:
![Screenshot from 202... | Bug,Help Wanted,Domain: Completion Lists | low | Critical |
888,474,173 | PowerToys | [Feature-Request] PT Run: custom order for Plugins | 
The new Plugin list with the ability to disable single elements is awesome. Thanks for this great implementation!
The feature would be even better if users could re-order the plugins to their own liking. ... | Idea-Enhancement,Run-Plugin Manager | low | Minor |
888,506,260 | flutter | RenderWrap.computeMaxIntrinsicWidth does not calculate spacing, resulting in improper line feeds. | code:
```dart
import 'package:flutter/material.dart';
void main() => runApp(MyApp());
class MyApp extends StatelessWidget {
static const String _title = 'Flutter Code Sample';
@override
Widget build(BuildContext context) {
var data = 'test test test test test test test test test test';
r... | framework,a: layout,has reproducible steps,P2,found in release: 3.7,found in release: 3.11,team-framework,triaged-framework | low | Minor |
888,690,260 | pytorch | TensorList upfront checks parameter types in argument parsing and throws an error unexpectedly | It has been discussed in the following PR https://github.com/pytorch/pytorch/pull/57529#discussion_r627059296 that in the process of argument parsing, unlike other lists (IntList, ScalarList etc) type checking wants to handle TensorLists parameters upfront whether TensorList is the correct parameter of signature or not... | triaged,module: codegen | low | Critical |
888,706,286 | godot | Editor tooltips should be constrained inside the editor window even at the bottom of the UI | <!-- Please search existing issues for potential duplicates before filing yours:
https://github.com/godotengine/godot/issues?q=is%3Aissue
-->
**Godot version:** 3.3.stable
<!-- Specify commit hash if using non-official build. -->
**OS/device including version:** Windows 10
<!-- Specify GPU model, drivers, a... | discussion,topic:editor,usability | low | Critical |
888,786,368 | TypeScript | Idea: Can we cache the result of `typeToTypeNode`? | A large percentage of declaration emit time (and a small percentage of check time) is spent in `typeToTypeNode` and tracing indicates that it is called on some types many times (hundreds or thousands in a sufficiently large project). I suspect, but have been able to demonstrate, that those requiring a `import()` quali... | Domain: Performance,Rescheduled | low | Minor |
888,794,848 | pytorch | Follow-up updates to inference mode python bindings | Issue tracks some things to do to follow up on #58045
- Move tests to separate file from `test_autograd`
- Move `inference_mode` context manager to its own file (from `grad_mode.py`
- Python bindings for `is_inference_tensor`
- Improved short description and example
- Link to tutorial | triaged,inference mode | low | Minor |
888,799,976 | TypeScript | Idea: Can declaration emit synthesize imports? | If an unimported type is referenced repeatedly, each reference will be qualified with the same `import("path")` expression. Cumulatively, these can significantly enlarge a declaration file.
Probably, this will be the first time we need to make non-local edits to a declaration file, which may substantially increase ... | Domain: Performance,Fix Available,Rescheduled | low | Minor |
888,815,993 | TypeScript | Idea: Can declaration emit synthesize type aliases? | Object types can be very large and in (e.g.) a react project that makes heavy use of type inference, it's not uncommon to have to refer to a type that is either not imported or is simply anonymous (e.g. a props type). When these appear in multiple places in a declaration file, they tend to expand it considerably. Per... | Suggestion,Domain: Performance,Experimentation Needed | low | Major |
888,851,850 | TypeScript | Idea: Distinguish between build-signature declarations and for-publication declarations | Some declaration files are intended for publication and they need to be readable and "work" (e.g. no imports from node_modules). Others are, from a user's perspective, basically temporary files produced by the build to make subsequent builds more incremental (i.e. faster). Right now, we produce both sets of declarati... | Domain: Performance | low | Minor |
888,890,694 | pytorch | TrilinearBackward takes 98.4% of total computational time, is this to be expected? | I made a new neural network yesterday that seems to be running very slow for some reason, so I used the profiler on it today and got the following. When just running it on the forward pass of the model I get the following:
```
----------------------------- ------------ ------------ ------------ ------------ --... | module: performance,module: nn,triaged | low | Critical |
889,252,148 | vscode | [api][commands] Open new window and a file | The `vscode.newWindow` command takes a `remoteAuthority` argument and opens a window on that authority. The Azure ML extension needs to open a window on a certain remote, then open a file on that remote. They do this by adding extra data to the remoteAuthority. This works but a downside is that you end up with extra en... | feature-request,api | low | Major |
889,499,355 | go | cmd/compile: allocate slices lazily | This is an alternative to #15735.
When a slice is constructed (`make([]foo, N)`), the Go runtime could avoid allocating the backing array until something actually tries to read or write to the slice. In most cases, the slice is used immediately, so this has little effect, but when used with a blocking implemention ... | Performance,NeedsInvestigation | low | Major |
889,522,515 | pytorch | torch.jit.trace memory leak | ## 🐛 Bug
torch.jit.trace has memory leak.
## To Reproduce
Run following code
import gc
import os
import sys
import torch
import psutil
nFeatures = 52
layer_neurons = [nFeatures,128,256,512,256,128,128,64,6... | oncall: jit | low | Critical |
889,532,806 | pytorch | Static graph training fails if forward is called multiple times before backward | ## 🐛 Bug
The accounting for how many times a grad will have its hook called is done in the backwards pass, namely in the autograd hook for the variable. However we only do this if `num_iterations_ == 1` so we implicitly have the assumption that every forward pass will do a backwards pass before calling the next for... | high priority,triage review,oncall: distributed,triaged,module: ddp | low | Critical |
889,543,873 | kubernetes | Ability to declare pods to be "recreated" on "failure" (instead of restarts) | The following is to track the issue raised originally in : https://github.com/kubernetes/kubernetes/issues/24725
That is specific to only the desired behaviour of "If Pod-XYZABC exits for any reason, do not restart it, provide me a new one" pattern.
#### What would you like to be added:
It has been suggested to ... | kind/feature,sig/apps,lifecycle/frozen,needs-triage | medium | Critical |
889,646,181 | pytorch | [FX] Revise PyTree support in FX to move pytree logic out of `self.code` | The idea is to move pytree-ification of a module into a separate pass, and have that pass return a tuple of `(flat_graph, reinstall_pytree_fn)`.
cc @ezyang @SherlockNoMad | triaged,module: fx | low | Minor |
889,662,573 | flutter | Rename DropdownButtonFormField.value to initialValue | ## Explaination
Like TextFormField, DropdownButtonFormField is "A convenience widget that makes a DropdownButton into a FormField."
FormField(s) maintains the current state internally, you only pass the initial value (like TextFormField)
& further more in the implementation it's just setting initialValue: value.
... | c: new feature,framework,f: material design,c: proposal,team-design,triaged-design | low | Minor |
889,691,680 | godot | Stuttering when a Skeleton first appears on the screen on GLES3 | **Godot version:** v3.3.stable.official
**OS/device including version:** macOS 11.3.1 / Intel Iris Graphics 6100 1536 MB / GLES3
**Issue description:**
Game stutters when a model with skeleton appears on the screen for the first time.
* This does not happen in GLES2
* This also happens in 3.2.3
**Steps ... | discussion,topic:rendering,topic:3d | low | Minor |
889,694,086 | TypeScript | Could there be a `This` type like the `Self` type in rust ? | # Suggestion
<!--
Please fill in each section completely. Thank you!
-->
## 🔍 Search Terms
This, Self, typeof this
## ✅ Viability Checklist
<!--
Suggestions that don't meet all these criteria are very, very unlikely to be accepted.
We always recommend reviewing the TypeScript design goals be... | Suggestion,Awaiting More Feedback | low | Major |
889,696,983 | pytorch | Generate special clear error messages for known common misuses in TorchScript | There are some common misuse patterns in TorchScript that we should issue clear error messages for instead of generating generic error that doesn't capture root cause of error.
Here are a few examples:
- Attempting to construct a `nn.Module` inside TorchScript. This currently errors out because TorchScript would ... | module: bootcamp,good first issue,triaged,OSS contribution wanted | medium | Critical |
889,726,417 | pytorch | Python API binding code generation does not need to pack TensorOptions for `xxx_like` generators | ## 🚀 Feature
Tensor generators like `zeros_like` have kernels that read tensor options like device/dtype from input tensor. Generated code for Python API bindings does not need to redundantly extract same information from input tensor at binding layer.
## Motivation
Normally extracting the tensor options early... | triaged,module: codegen | low | Minor |
889,729,174 | flutter | AlignmentGeometryTween should extend Tween<AlignmentGeometry> (not nullable) instead of Tween<AlignmentGeometry?> (nullable) | The class `AlignmentGeometryTween` extends from `Tween<AlignmentGeometry?>`, this means that when creating an animation out of it, the animation will be an `Animation<AlignmentGeometry?>`. (note how AlignmentGeometry can be null)
https://github.com/flutter/flutter/blob/e7b7ebc066c1b2a5aa5c19f8961307427e0142a6/packa... | framework,a: animation,f: material design,has reproducible steps,a: null-safety,P2,found in release: 2.3,team-design,triaged-design | low | Critical |
889,756,375 | flutter | `LateError: LateInitializationError: Local 'built' has not been initialized.` | Hello,
I am using latest flutter and flutter doctor shows no issues.
I see (live app) many errors like this: `LateError: LateInitializationError: Local 'built' has not been initialized.`
I am not sure is SilverAppBar is related but I saw those errors mostly when SIlverAppBar is involved.
I do not know how to so... | c: crash,framework,a: null-safety,P2,team-framework,triaged-framework | low | Critical |
889,801,738 | pytorch | Save checkpoint error | when I save the lr_scheduler.state_dict() to checkpoint, torch.save will report following error
```
Traceback (most recent call last):
File "pytorch_imagenet_resnet50.py", line 526, in <module>
save_checkpoint(epoch)
File "pytorch_imagenet_resnet50.py", line 311, in save_checkpoint
torch.save(state,... | module: optimizer,module: serialization,triaged | low | Critical |
889,812,624 | ant-design | Is it possible to set ant table column width in percantage (px is work) with scroll={{ x: "max-content" }} and empty data? Is it bug or not? | - [ ] I have searched the [issues](https://github.com/ant-design/ant-design/issues) of this repository and believe that this is not a duplicate.
### Reproduction link
[](https://codesandbox.io/s/basic-usage-antd4155-forked-bg2gs)
### S... | help wanted,Inactive | low | Critical |
889,821,116 | pytorch | the same input with different batchsize got different precision output | ## Issue description
Provide a short description.
## Code example
I have an Transform model
when i do inference with an input:
`att_out, ctc_out, state1 = encode_model(inputs, length, outlength, num_idb, state)`
and then, I got a batch input by followings:
` batch = int(sys.argv[1])
inputs = to... | module: nn,triaged,module: numerical-reproducibility,module: batching | low | Critical |
889,974,223 | rust | Suggest restricting a type parameter with clone when attempting to clone out a ref | Not entirely sure about the title, but the code is simple enough:
```rust
pub struct Wrapper<T>(Option<T>);
impl<T> Wrapper<T> {
pub fn fail(&self) -> Option<T> {
self.0.as_ref().map(|x| x.clone())
}
}
```
Yields this error message:
```
error[E0308]: mismatched types
--> src/lib.rs:5:... | A-diagnostics,T-compiler | low | Critical |
890,012,124 | youtube-dl | Support async/await in progress hooks | <!--
######################################################################
WARNING!
IGNORING THE FOLLOWING TEMPLATE WILL RESULT IN ISSUE CLOSED AS INCOMPLETE
######################################################################
-->
## Checklist
<!--
Carefully read and work through this check lis... | request | low | Critical |
890,037,456 | flutter | Trackpad scrolling is not the same between Flutter Web and Flutter Windows | OLD ISSUE:
<details>
- Trackpad scrolling produces a delta of zero in a Windows Window, but not Flutter web on Windows.
- onPointerSignal in the Listener widget does NOT fire when any other key is pressed. This works in Flutter web, but not Windows. It fires extremely inconsistently on Windows.
Here's a dartpad r... | framework,engine,platform-windows,a: desktop,has reproducible steps,P2,found in release: 2.0,found in release: 2.3,team-windows,triaged-windows | low | Major |
890,057,669 | material-ui | [List] ListItemIcon color is too strict | <!-- Checked checkbox should look like this: [x] -->
- [x] The issue is present in the latest release.
- [x] I have searched the [issues](https://github.com/mui-org/material-ui/issues) of this repository and believe that this is not a duplicate.
## Current Behavior 😯
![image](https://user-images.githubuserco... | new feature,component: list | low | Major |
890,057,872 | TypeScript | "Extract to function": parameter type missing when extracting code that depends on `node_modules` | # Bug Report
<!--
Please fill in each section completely. Thank you!
-->
### 🔎 Search Terms
<!--
What search terms did you use when trying to find an existing bug report?
List them here so people in the future can find this one more easily.
-->
code action quick fix module scope extract function... | Bug,Help Wanted,Domain: Refactorings | low | Critical |
890,075,250 | pytorch | Segmentation fault need help | ## 🐛 Bug
<!-- A clear and concise description of what the bug is. -->
the codes works pretty well in CPU mode (tested in my macbook pro), and get Segmentation fault (core dumped) in GPU mode, which is my pc at home.
## To Reproduce
```
==============Model Training===========
{'batch_size': 64, 'buffer_si... | needs reproduction,module: cuda,triaged | low | Critical |
890,081,088 | pytorch | scatter_add_ 6000-times slower with int64 compared to int32 | ## 🐛 Bug
`scatter_add_` is unreasonably slow on GPU for the `torch.long` data type. I understand this might be closely related to the hardware layout of GPUs, however, not getting a user warning or a mention in the docs is extremely irritating.
## To Reproduce
Steps to reproduce the behavior:
1. Run the be... | module: performance,module: docs,module: cuda,triaged,module: scatter & gather ops | low | Critical |
890,119,939 | PowerToys | Audible Feedback for PowerToys Run | ### Provide a description of the new feature / enhancement
When invoking PowerToys Run it would be nice to have audible feedback. The run box is just a black rectangle which can sometimes be lost on a busy screen and can take a while to display. | Idea-Enhancement,Help Wanted,Product-PowerToys Run,Cost-Small | low | Minor |
890,122,534 | rust | Specialization on associated type | Specializing on an associated type doesn't yet work.
```rust
#![feature(specialization)]
trait Tr {
type X;
fn f(&self) -> Self::X;
}
// Works
// impl<T> Tr for T {
// type X = i8;
// fn f(&self) -> Self::X { 1_i8 }
// }
// Works
// impl<T> Tr for T {
// type X = i8;
//
... | A-diagnostics,T-compiler,A-specialization,C-bug,F-specialization,D-confusing | low | Critical |
890,141,237 | rust | simd_shuffle, simd_insert, simd_extract should use const generics | `simd_shuffle` requires its last parameter, the array defining the shuffling, to be a constant:
https://github.com/rust-lang/rust/blob/487e27350a16e35f77ca471f368be3da28dac27c/compiler/rustc_codegen_llvm/src/intrinsic.rs#L888
Currently (after https://github.com/rust-lang/rust/pull/85110), this is enforced by an ad-... | T-compiler,A-SIMD,A-const-generics,A-intrinsics | medium | Major |
890,172,789 | pytorch | [docs] Strange torch.unique function signature | `torch.unique(*args, **kwargs)` - haven't seen any other functions with signatures not listing the arguments inline
UPD: still the case, https://pytorch.org/docs/master/generated/torch.unique.html#torch.... | module: docs,module: bootcamp,triaged | low | Minor |
890,180,444 | ant-design | Progress Step size | - [x] I have searched the [issues](https://github.com/ant-design/ant-design/issues) of this repository and believe that this is not a duplicate.
### What problem does this feature solve?
When "steps" are enabled, the progress bar uses a fixed with of 14px per step. Other progress bars have a width 100% (by default... | help wanted,Inactive | low | Major |
890,218,441 | puppeteer | FileChooser.accept doesn't work on file dialogs not associated with an input element | Here is my code
[nodeJs]
```
const puppeteer = require('puppeteer');
class App {
private static ROOT_PATH:string = "D:/WWW/test_puppeteer_file_access/"
private static CLIENT_PATH:string = "http://127.0.0.1/edsa-LocalHost_D/test_puppeteer_file_access/src/index.html";
private page:any;
co... | bug,upstream,confirmed,P2 | low | Major |
890,251,495 | node | stream: issues with Duplex streams (especially not closing/destroying properly) | First, the code:
```js
'use strict';
const { Duplex } = require('stream');
class Foo extends Duplex {
constructor(isClient) {
super({ autoDestroy: true });
this.isClient = isClient;
}
_read(n) {}
_write(data, encoding, cb) {
this.emit('bar');
cb();
}
_destroy(err, cb) {
... | stream | low | Critical |
890,264,657 | pytorch | "uncorrectable NVLink error" making the tests fail | ## 🐛 Bug
Trying to build PyTorch 1.8.1 on our latest A100 cluster fails during the tests. The failing tests are
- test_nccl_errors_blocking_abort
- test_nccl_errors_blocking_nonzero_exit
- test_nccl_errors_blocking_sigkill
- test_nccl_errors_blocking_sigterm
- test_nccl_blocking_wait_with_barrier
- test_ncc... | oncall: distributed,module: cuda,triaged,module: nccl | low | Critical |
890,328,815 | flutter | [Macos][Desktop][Feature Request]: Add FlutterViewController's initWithEngine method | try upgrade flutter
try change channel to dev
deleted flutter/bin/cache/...FlutterMacOS and redownload
the same... | c: new feature,engine,platform-mac,c: proposal,a: desktop,P2,team-macos,triaged-macos | low | Major |
890,373,916 | pytorch | Allow `Union` as a `Dict` key | ## 🚀 Feature
If we have `Union[T1, T2, ... TN]` where each internal type `TK` is an allowable dict key, then we should also allow that Union as a dict key. If one or more of the types in the Union is not an allowable Dict key, we shouldn't allow that Union as a dict key.
The difficulty here comes from making schem... | triaged,TSRootCause:TypeRefinement | low | Minor |
890,374,449 | go | cmd/compile: failures with "wait: no child processes" on solaris-amd64-oraclerel since 2021-04-02 | ```
#!watchflakes
post <- goos == "solaris" && log ~ `wait: no child processes`
```
```
##### ../test
go tool compile: wait: no child processes
go tool dist: FAILED: go list -gcflags=all= -ldflags=all= -f={{if .Stale}} STALE {{.ImportPath}}: {{.StaleReason}}{{end}} std: exit status 1
```
[2021-05-12T14:4... | OS-Solaris,NeedsInvestigation,compiler/runtime | medium | Critical |
890,383,861 | PowerToys | Toggle on old Photoviewer | I really miss the old Photviewer from Windows past. I know I can use it with a manual registry entry on Windows 10, but it would be so much nicer if it was just a resident Powertoy and be able to toggle it on with a simple push of a button. | Idea-New PowerToy,Product-Tweak UI Design | low | Minor |
890,384,767 | flutter | [web] `showDatePicker` on web shows an error when date selected and uncaught exceptions is turned ON | ## Steps to Reproduce
Create a default flutter project. Replace the code in the `main.dart` with the below:
```dart
import 'package:flutter/material.dart';
void main() => runApp(MyApp());
class MyApp extends StatelessWidget {
@override
Widget build(BuildContext context) {
return MaterialApp(
... | engine,f: date/time picker,a: debugging,platform-web,a: error message,has reproducible steps,P2,found in release: 2.3,team-web,triaged-web | low | Critical |
890,399,462 | pytorch | Error while converting model for NNAPI "Unsupported node kind ('quantized::batch_norm2d')" | ## 🐛 Bug
I'm converting the Unet model from (torch-hub)[https://pytorch.org/hub/mateuszbuda_brain-segmentation-pytorch_unet/] to run on Android for a benchmark. Error happens after I've quantized the model and converted to use `NHWC`.
## To Reproduce
Steps to reproduce the behavior:
```python
model = tor... | triaged,module: android,oncall: mobile | low | Critical |
890,404,864 | pytorch | Mobile Android: Could not run 'aten::quantize_per_tensor' with arguments from the 'Vulkan' backend | ## 🐛 Bug
When running a model with vulkan on android (converted using the mobile recipes on PyTorch Mobile page), this error happens on forward pass. Works fine when inferencing on CPU.
## To Reproduce
### Mobile Code:
```Java
try {
bitmap = ((BitmapDrawable)imageView.getDrawable()).get... | triaged,module: android,oncall: mobile | low | Critical |
890,429,515 | node | Stream finished does not always work with http incoming message | * **Version**: v15.12.0
* **Platform**: Darwin Kernel Version 20.3.0
* **Subsystem**: http
### What steps will reproduce the bug?
The `stream.finished` never resolves or rejects when applied onto a destroyed incoming message like in example below. It finishes for example when applied on a destroyed file stream.... | stream | medium | Critical |
890,444,449 | pytorch | torch.split_with_sizes is not documented | 
The function is available and tested. Also the name does not start with a leading underscore so I assume it is meant to be public. Looks like this was just a miss.
https://gi... | module: docs,triaged | low | Major |
890,453,307 | vscode | Git conflict resolution Accept Current Change "ignores" addition of a single empty line | <!-- ⚠️⚠️ Do Not Delete This! bug_report_template ⚠️⚠️ -->
<!-- Please read our Rules of Conduct: https://opensource.microsoft.com/codeofconduct/ -->
<!-- 🕮 Read our guide about submitting issues: https://github.com/microsoft/vscode/wiki/Submitting-Bugs-and-Suggestions -->
<!-- 🔎 Search existing issues to avoid cr... | bug,merge-conflict | low | Critical |
890,467,707 | flutter | implement flutter precache --dry-run | When passed the flag `--dry-run`, instead of downloading remote artifacts, the tool would just print out the remote path it would have downloaded, and perhaps the relative path in the cache it would have downloaded to (maybe support `--machine` so this could be structured JSON)
This would assist release tooling that... | team,P3,team-tool,triaged-tool | low | Minor |
890,518,578 | terminal | Add settings import to command line | <!--
🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
I ACKNOWLEDGE THE FOLLOWING BEFORE PROCEEDING:
1. If I delete this entire template and go my own path, the core team may close my issue without further explanation or engagement.
2. If I list multiple bugs/concerns in this one issue, the core team may close my issue without further expl... | Issue-Feature,Needs-Tag-Fix,Area-Commandline | low | Critical |
890,520,380 | pytorch | [FX] record_function tracing in FX | ```
import torch
import torch.fx
class Foo(torch.nn.Module):
def forward(self, x):
with torch.profiler.record_function('my_op'):
return torch.relu(x)
traced = torch.fx.symbolic_trace(Foo())
print(traced.code)
"""
def forward(self, x):
relu = torch.relu(x); x = None
retur... | triaged,module: fx | low | Minor |
890,531,609 | neovim | vim.list_extend vs. vim.tbl_extend | `vim.list_extend` and `vim.tbl_extend` are inconsistent -- the latter returns a copy, while the former mutates its argument.
Is it too late to "fix" / change that? I'd love a `list_extend` that returns a copy instead, it seems to be a common need for me when writing Lua code, and then if versions with mutation are d... | enhancement,lua | low | Major |
890,539,618 | PowerToys | Monitor file changes in Markdown preview, update when file is changed. | ### Microsoft PowerToys version
0.37.0
### Running as admin
- [X] Yes
### Area(s) with issue?
MD Preview
### Steps to reproduce
1. Open Windows File Explorer.
2. Click on Markdown file and view it in the preview pane.
3. Open the Markdown file in VSCode.
4. Edit the Markdown file in VSCode.
5. Save the Markd... | Idea-Enhancement,Product-File Explorer | low | Minor |
890,610,882 | rust | Doc example with feature dependency does not compile | A doc example which depends on a feature does not compile if the entire example is conditionally excluded.
Having the example conditionally compiled with a feature does not work if the feature is not enabled:
```rust
/// ```
/// #![cfg(feature = "foo")]
/// assert_eq!(1, 1);
/// ```
pub struct Foo;
```
F... | T-rustdoc,A-doctests,C-discussion | low | Critical |
890,611,357 | flutter | [in_app_purchase] iOS Always prompting for authentication on every purchase in sandbox mode | On IOS only after every purchase it prompts the user for their apple id and password. This behavior exists after deploying to test flight as well.
i am doing to make the purchase. The purchase works its just asking me for my apple id and password its doing the same for all my beta testers as well.
```dart
... | platform-ios,a: quality,p: in_app_purchase,package,has reproducible steps,P2,found in release: 2.0,team-ios,triaged-ios | low | Critical |
890,631,707 | tensorflow | Compilation fails on Ubuntu 20.04 when using TensorRT 8. | **System information**
- OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 20.04
- Mobile device (e.g. iPhone 8, Pixel 2, Samsung Galaxy) if the issue happens on mobile device: N/A
- TensorFlow installed from (source or binary): binary
- TensorFlow version: 2.4.1, 2.5, etc
- Python version: 3.8
- In... | stat:awaiting tensorflower,type:bug,subtype: ubuntu/linux,comp:gpu:tensorrt,TF 2.5 | medium | Critical |
890,631,765 | youtube-dl | loco.gg | <!--
######################################################################
WARNING!
IGNORING THE FOLLOWING TEMPLATE WILL RESULT IN ISSUE CLOSED AS INCOMPLETE
######################################################################
-->
## Checklist
<!--
Carefully read and work through this check lis... | site-support-request | low | Critical |
890,636,270 | rust | Stack overflow/infinite recursion in `rustc_middle::ty::fold::TypeFoldable::fold_with` | https://gitlab.com/veloren/veloren/-/commit/a8b267387cf4fdc0d537668defa20e1eeecf701e is the version of Veloren's codebase that triggers the bug.
It initially manifested as a SIGSEGV, looking at it under gdb seems to indicate that it's infinite recursion.
Rustc/cargo versions:
```
$ cargo --version
cargo 1.52.0-n... | A-type-system,I-crash,P-medium,T-compiler,C-bug,T-types | low | Critical |
890,639,945 | tauri | Feature request: JS API throws Error objects instead of strings | **Is your feature request related to a problem? Please describe.**
As a web developer who works primarily in with HTML/CSS/JS, all of my app code is currently JS which calls the tauri APIs via its JS bindings when appropriate. When an expected exception occurs (for example, trying to read a file that doesn't exist), I... | scope: core,type: feature request,priority: 3 low,type: breaking change | low | Critical |
890,654,356 | pytorch | `torch.autograd.Function` subclasses *sometimes* throw away custom subclasses | ## 🐛 Bug
`apply`ing a torch.autograd.Function usually returns the same `torch.Tensor` subclass as the input, but not always. This seems to happen when an input `requires_grad` and potentially some other requirements.
## To Reproduce
Steps to reproduce the behavior:
```python
import torch
import torch.aut... | module: autograd,triaged,actionable | low | Critical |
890,659,068 | TypeScript | Infer type isn't contextual typed by type bounds | # Bug Report
Infer type isn't contextual typed by type bounds
```ts
type Foo<T> = T extends string ? Record<T, number> : never;
type Foo2<T extends {a:string}>= T extends {a: infer A} ? Record<A, number>:never; //error here, A should extends string already, but seems not
```
<!--
Please fill in each section... | Bug | low | Critical |
890,672,297 | flutter | documentation: Upgrading pre 1.12 Android projects: step 1b is missing Kotlin sample code | #50004 seems to have the right sample code. It'd be nice to include it so that readers don't have to try porting the sample Java code to Kotlin themselves. | platform-android,d: wiki,P3,team-android,triaged-android | low | Minor |
890,672,983 | rust | Error message contains `$crate` instead of crate name | Since https://github.com/rust-lang/rust/pull/82069 was merged, the 'this error originates in macro `<name>`) messages sometimes contain `$crate`, instead of the actual crate name.
For example:
https://github.com/rust-lang/rust/blob/9daf546b77dbeab7754a80d7336cd8d00c6746e4/src/test/ui/bound-suggestions.stderr#L8
... | A-diagnostics,A-macros,T-compiler,C-bug | low | Critical |
890,699,621 | flutter | TextRange.isValid can't work | ```
$ flutter doctor
Doctor summary (to see all details, run flutter doctor -v):
[✓] Flutter (Channel stable, 2.0.4, on Mac OS X 10.15.5 19F101 darwin-x64, locale zh-Hans-CN)
[✓] Android toolchain - develop for Android devices (Android SDK version 30.0.3)
[✓] Xcode - develop for iOS and macOS
[✓] Chrome - develo... | a: text input,framework,a: internationalization,has reproducible steps,P2,found in release: 2.3,team-framework,triaged-framework | low | Major |
890,728,222 | pytorch | Torch.nn.DataParallel training model, the output of the model becomes list type, the number of lists is batch size | Using Torch.nn.DataParallel to train model, the output of the model becomes list type, the number of lists is batch size.
The first epoch training output is normal,the output types are as follows：
model= torch.nn.DataParallel(model,device_ids=[0,1]).to('cuda')
out = model(data) # out size: [batch, c,h,w]
But for ... | triaged,module: data parallel | low | Minor |
Subsets and Splits
GitHub Issues Containing Next.js References
Filters training data to find examples mentioning "next.js", providing basic keyword search capability but offering limited analytical value beyond simple retrieval.