Add a vision transformer example (dino-v2). (#502)

* Add a vision transformer example (dino-v2).

* Add some documentation + test.

* CI fix.

* Another fix (still unable to replicate the errors locally :( )

candle-book/src/guide/hello_world.md

@@ -147,7 +147,7 @@ And rewrite our examples using it
 # extern crate candle_core;
 # extern crate candle_nn;
 use candle_core::{DType, Device, Result, Tensor};
-use candle_nn::Linear;
+use candle_nn::{Linear, Module};
 
 struct Model {
     first: Linear,

candle-book/src/inference/hub.md

@@ -58,7 +58,7 @@ Now that we have our weights, we can use them in our bert architecture:
 # 
 # let weights = repo.get("model.safetensors").unwrap();
 use candle_core::{Device, Tensor, DType};
-use candle_nn::Linear;
+use candle_nn::{Linear, Module};
 
 let weights = candle_core::safetensors::load(weights, &Device::Cpu).unwrap();