diff --git a/src/convertor.rs b/src/convertor.rs
index 4e7c771..b28ecb1 100644
--- a/src/convertor.rs
+++ b/src/convertor.rs
@@ -71,7 +71,7 @@ fn read_collection() -> Result<Vec<RevlogEntry>> {
 }
 
 fn group_by_cid(revlogs: Vec<RevlogEntry>) -> Vec<Vec<RevlogEntry>> {
-    let mut grouped: HashMap<i64, Vec<RevlogEntry>> = HashMap::new();
+    let mut grouped = HashMap::new();
     for revlog in revlogs {
         grouped
             .entry(revlog.cid)
diff --git a/src/dataset.rs b/src/dataset.rs
index 1ec77c9..8d90368 100644
--- a/src/dataset.rs
+++ b/src/dataset.rs
@@ -64,7 +64,7 @@ impl<B: Backend> Batcher<FSRSItem, FSRSBatch<B>> for FSRSBatcher<B> {
         let (time_histories, rating_histories) = items
             .iter()
             .map(|item| {
-                let (mut delta_t, mut rating): (Vec<i32>, Vec<i32>) =
+                let (mut delta_t, mut rating): (Vec<_>, Vec<_>) =
                     item.history().map(|r| (r.delta_t, r.rating)).unzip();
                 delta_t.resize(pad_size, 0);
                 rating.resize(pad_size, 0);
@@ -187,7 +187,7 @@ fn test_batcher() {
     use burn_ndarray::NdArrayDevice;
     type Backend = NdArrayBackend<f32>;
     let device = NdArrayDevice::Cpu;
-    let batcher: FSRSBatcher<Backend> = FSRSBatcher::<Backend>::new(device);
+    let batcher = FSRSBatcher::<Backend>::new(device);
     let items = vec![
         FSRSItem {
             reviews: vec![
diff --git a/src/training.rs b/src/training.rs
index dd9b8bf..efc6e56 100644
--- a/src/training.rs
+++ b/src/training.rs
@@ -1,3 +1,5 @@
+use std::path::Path;
+
 use crate::cosine_annealing::CosineAnnealingLR;
 use crate::dataset::{FSRSBatch, FSRSBatcher, FSRSDataset};
 use crate::model::{Model, ModelConfig};
@@ -107,7 +109,12 @@ pub fn train<B: ADBackend<FloatElem = f32>>(
 ) {
     std::fs::create_dir_all(artifact_dir).ok();
     config
-        .save(&format!("{artifact_dir}/config.json"))
+        .save(
+            Path::new(artifact_dir)
+                .join("config.json")
+                .to_str()
+                .unwrap(),
+        )
         .expect("Save without error");
 
     B::seed(config.seed);
@@ -153,13 +160,18 @@ pub fn train<B: ADBackend<FloatElem = f32>>(
     info!("clipped weights: {}", &model_trained.w.val());
 
     config
-        .save(format!("{ARTIFACT_DIR}/config.json").as_str())
+        .save(
+            Path::new(ARTIFACT_DIR)
+                .join("config.json")
+                .to_str()
+                .unwrap(),
+        )
         .unwrap();
 
     PrettyJsonFileRecorder::<FullPrecisionSettings>::new()
         .record(
             model_trained.into_record(),
-            format!("{ARTIFACT_DIR}/model").into(),
+            Path::new(ARTIFACT_DIR).join("model"),
         )
         .expect("Failed to save trained model");
 }