mirror of
https://github.com/huggingface/candle.git
synced 2025-06-16 18:48:51 +00:00
90374097dc
* Add a cudnn feature to be used for conv2d. * Allocate the proper workspace. * Only create a single cudnn handle per cuda device. * Proper cudnn usage. * Bugfix.
18 lines
438 B
Rust
18 lines
438 B
Rust
#[cfg(feature = "accelerate")]
|
|
extern crate accelerate_src;
|
|
|
|
#[cfg(feature = "mkl")]
|
|
extern crate intel_mkl_src;
|
|
|
|
use anyhow::Result;
|
|
use candle_core::{Device, Tensor};
|
|
|
|
fn main() -> Result<()> {
|
|
let device = Device::new_cuda(0)?;
|
|
let t = Tensor::randn(0f32, 1f32, (2, 4, 96, 96), &device)?;
|
|
let w = Tensor::randn(0f32, 1f32, (320, 4, 3, 3), &device)?;
|
|
let res = t.conv2d(&w, 1, 1)?;
|
|
println!("{res:?}");
|
|
Ok(())
|
|
}
|