WU Chen

I have an example: ``` struct Angle {} struct Point {} struct ArcSize {} struct ArcSweep {} pub trait Path { /// Adds an arc arc_to( Self, Double, // rx...

> We'll do some digging on our side (the Vite extension) to see if we can work around this somehow but I can't make any guarantees :/ @philipp-spiess Thank you....

I tried printing the tensor converted from the input f32 slice in the console. It seems that the tensor's data has been corrupted from the very beginning. 

Is this bug caused by `Pool2dEagerKernel`? @nathanielsimard @louisfd  ```` [START_KERNEL_COMPILATION] name: burn_jit::kernel::pool::pool2d_shader::Pool2dEagerKernel< burn_jit::kernel::pool::max_pool2d::MaxPool< f32, >, cubecl_wgpu::runtime::WgpuRuntime, f32, > cube_dim: (16, 16, 1) shared_memory: 0 bytes source: ``` @group(0) @binding(0)...

This bug is caused by the limited default precision of Rust's display of f32 values. It should be an easy fix. I changed the code from https://github.com/tracel-ai/cubecl/blob/cfe0b0204380cbd0931f478194a053a6ac35d1cb/crates/cubecl-wgpu/src/compiler/wgsl/base.rs#L262-L263 : ```rust FloatKind::F32...

In fact, I've tried all the release versions of Burn since 10.0.0 on my device, but the wgpu backend in the `image-classification-example` hasn't worked. However, after the small modifications I...

@Jonarod You also need to disable the use of the autotune feature for burn-wgpu. As for Candle, I don't know how to make it work either.

@laggui This is because I saw in the browser console that -3.40282347E+38f32 was rounded to -340282350000000000000000000000000000000f, which caused a WGSL compilation error, so I believe this is an issue with...

It looks like this issue has been resolved in [cubecl 130](https://github.com/tracel-ai/cubecl/pull/130).