Remove some unused macros. (#2618)

* Remove some unused macros.

* More unused fixes.

candle-examples/examples/reinforcement-learning/ddpg.rs

@@ -1,5 +1,4 @@
 use std::collections::VecDeque;
-use std::fmt::Display;
 
 use candle::{DType, Device, Error, Module, Result, Tensor, Var};
 use candle_nn::{
@@ -167,6 +166,7 @@ fn track(
     Ok(())
 }
 
+#[allow(unused)]
 struct Actor<'a> {
     varmap: VarMap,
     vb: VarBuilder<'a>,
@@ -211,7 +211,7 @@ impl Actor<'_> {
         let target_network = make_network("target-actor")?;
 
         // this sets the two networks to be equal to each other using tau = 1.0
-        track(&mut varmap, &vb, "target-actor", "actor", &dims, 1.0);
+        track(&mut varmap, &vb, "target-actor", "actor", &dims, 1.0)?;
 
         Ok(Self {
             varmap,
@@ -244,6 +244,7 @@ impl Actor<'_> {
     }
 }
 
+#[allow(unused)]
 struct Critic<'a> {
     varmap: VarMap,
     vb: VarBuilder<'a>,
@@ -287,7 +288,7 @@ impl Critic<'_> {
         let target_network = make_network("target-critic")?;
 
         // this sets the two networks to be equal to each other using tau = 1.0
-        track(&mut varmap, &vb, "target-critic", "critic", &dims, 1.0);
+        track(&mut varmap, &vb, "target-critic", "critic", &dims, 1.0)?;
 
         Ok(Self {
             varmap,
@@ -322,6 +323,7 @@ impl Critic<'_> {
     }
 }
 
+#[allow(unused)]
 #[allow(clippy::upper_case_acronyms)]
 pub struct DDPG<'a> {
     actor: Actor<'a>,

candle-examples/examples/reinforcement-learning/gym_env.rs

@@ -1,4 +1,3 @@
-#![allow(unused)]
 //! Wrappers around the Python API of Gymnasium (the new version of OpenAI gym)
 use candle::{Device, Result, Tensor};
 use pyo3::prelude::*;

candle-examples/examples/reinforcement-learning/main.rs

@@ -1,5 +1,3 @@
-#![allow(unused)]
-
 #[cfg(feature = "mkl")]
 extern crate intel_mkl_src;
 

candle-examples/examples/reinforcement-learning/policy_gradient.rs

@@ -14,7 +14,7 @@ fn new_model(
 ) -> Result<(impl Module, VarMap)> {
     let input_size = input_shape.iter().product();
 
-    let mut varmap = VarMap::new();
+    let varmap = VarMap::new();
     let var_builder = VarBuilder::from_varmap(&varmap, dtype, device);
 
     let model = seq()

candle-examples/examples/reinforcement-learning/vec_gym_env.rs

@@ -1,9 +1,8 @@
-#![allow(unused)]
 //! Vectorized version of the gym environment.
 use candle::{DType, Device, Result, Tensor};
 use pyo3::prelude::*;
-use pyo3::types::PyDict;
 
+#[allow(unused)]
 #[derive(Debug)]
 pub struct Step {
     pub obs: Tensor,
@@ -11,6 +10,7 @@ pub struct Step {
     pub is_done: Tensor,
 }
 
+#[allow(unused)]
 pub struct VecGymEnv {
     env: PyObject,
     action_space: usize,
@@ -21,6 +21,7 @@ fn w(res: PyErr) -> candle::Error {
     candle::Error::wrap(res)
 }
 
+#[allow(unused)]
 impl VecGymEnv {
     pub fn new(name: &str, img_dir: Option<&str>, nprocesses: usize) -> Result<VecGymEnv> {
         Python::with_gil(|py| {