diff --git a/Cargo.toml b/Cargo.toml
index fa85db3..25d6bd9 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -70,7 +70,7 @@ features = ["doc-only"]
 
 [dependencies]
 rust_tokenizers = "~7.0.2"
-tch = "~0.9.0"
+tch = "~0.10.1"
 serde_json = "1.0.82"
 serde = { version = "1.0.140", features = ["derive"] }
 ordered-float = "3.0.0"
@@ -88,6 +88,6 @@ anyhow = "1.0.58"
 csv = "1.1.6"
 criterion = "0.3.6"
 tokio = { version = "1.20.0", features = ["sync", "rt-multi-thread", "macros"] }
-torch-sys = "0.9.0"
+torch-sys = "0.10.0"
 tempfile = "3.3.0"
 itertools = "0.10.3"
diff --git a/README.md b/README.md
index a860c02..4b41100 100644
--- a/README.md
+++ b/README.md
@@ -76,8 +76,8 @@ This cache location defaults to `~/.cache/.rustbert`, but can be changed by sett
 
 ### Manual installation (recommended)
 
-1. Download `libtorch` from https://pytorch.org/get-started/locally/. This package requires `v1.13.0`: if this version is no longer available on the "get started" page,
-the file should be accessible by modifying the target link, for example `https://download.pytorch.org/libtorch/cu117/libtorch-cxx11-abi-shared-with-deps-1.13.0%2Bcu117.zip` for a Linux version with CUDA11.
+1. Download `libtorch` from https://pytorch.org/get-started/locally/. This package requires `v1.13.1`: if this version is no longer available on the "get started" page,
+the file should be accessible by modifying the target link, for example `https://download.pytorch.org/libtorch/cu117/libtorch-cxx11-abi-shared-with-deps-1.13.1%2Bcu117.zip` for a Linux version with CUDA11. **NOTE:** When using `rust-bert` as dependency from [crates.io](https://crates.io), please check the required `LIBTORCH` on the published package [readme](https://crates.io/crates/rust-bert) as it may differ from the version documented here (applying to the current repository version).
 2. Extract the library to a location of your choice
 3. Set the following environment variables
 ##### Linux:
diff --git a/examples/natural_language_inference_deberta.rs b/examples/natural_language_inference_deberta.rs
index d63fa9e..e779a07 100644
--- a/examples/natural_language_inference_deberta.rs
+++ b/examples/natural_language_inference_deberta.rs
@@ -38,7 +38,7 @@ fn main() -> anyhow::Result<()> {
         false,
     )?;
     let config = DebertaConfig::from_file(config_path);
-    let model = DebertaForSequenceClassification::new(&vs.root(), &config);
+    let model = DebertaForSequenceClassification::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
diff --git a/src/bart/bart_model.rs b/src/bart/bart_model.rs
index edb4f69..2b9eb32 100644
--- a/src/bart/bart_model.rs
+++ b/src/bart/bart_model.rs
@@ -1101,7 +1101,7 @@ impl BartGenerator {
         generate_config.validate();
         let mut var_store = nn::VarStore::new(device);
         let config = BartConfig::from_file(config_path);
-        let model = BartForConditionalGeneration::new(&var_store.root(), &config);
+        let model = BartForConditionalGeneration::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = Some(config.bos_token_id.unwrap_or(0));
@@ -1131,7 +1131,7 @@ impl BartGenerator {
     }
 
     fn force_token_id_generation(&self, scores: &mut Tensor, token_ids: &[i64]) {
-        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size() as i64)
+        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size())
             .filter(|pos| !token_ids.contains(pos))
             .collect();
         let impossible_tokens = Tensor::of_slice(&impossible_tokens).to_device(scores.device());
@@ -1337,6 +1337,6 @@ mod test {
         let vs = tch::nn::VarStore::new(device);
         let config = BartConfig::from_file(config_path);
 
-        let _: Box<dyn Send> = Box::new(BartModel::new(&vs.root(), &config));
+        let _: Box<dyn Send> = Box::new(BartModel::new(vs.root(), &config));
     }
 }
diff --git a/src/bert/bert_model.rs b/src/bert/bert_model.rs
index 6868c1b..6b001bf 100644
--- a/src/bert/bert_model.rs
+++ b/src/bert/bert_model.rs
@@ -24,7 +24,7 @@ use crate::{Config, RustBertError};
 use serde::{Deserialize, Serialize};
 use std::borrow::Borrow;
 use std::collections::HashMap;
-use tch::nn::Init;
+use tch::nn::init::DEFAULT_KAIMING_UNIFORM;
 use tch::{nn, Kind, Tensor};
 
 /// # BERT Pretrained model weight files
@@ -507,7 +507,7 @@ impl BertLMPredictionHead {
             config.vocab_size,
             Default::default(),
         );
-        let bias = p.var("bias", &[config.vocab_size], Init::KaimingUniform);
+        let bias = p.var("bias", &[config.vocab_size], DEFAULT_KAIMING_UNIFORM);
 
         BertLMPredictionHead {
             transform,
@@ -1301,9 +1301,9 @@ mod test {
 
         //    Set-up masked LM model
         let device = Device::cuda_if_available();
-        let vs = tch::nn::VarStore::new(device);
+        let vs = nn::VarStore::new(device);
         let config = BertConfig::from_file(config_path);
 
-        let _: Box<dyn Send> = Box::new(BertModel::<BertEmbeddings>::new(&vs.root(), &config));
+        let _: Box<dyn Send> = Box::new(BertModel::<BertEmbeddings>::new(vs.root(), &config));
     }
 }
diff --git a/src/common/linear.rs b/src/common/linear.rs
index 48ad6c0..83cac42 100644
--- a/src/common/linear.rs
+++ b/src/common/linear.rs
@@ -11,6 +11,7 @@
 // limitations under the License.
 
 use std::borrow::Borrow;
+use tch::nn::init::DEFAULT_KAIMING_UNIFORM;
 use tch::nn::{Init, Module, Path};
 use tch::Tensor;
 
@@ -22,7 +23,7 @@ pub struct LinearNoBiasConfig {
 impl Default for LinearNoBiasConfig {
     fn default() -> Self {
         LinearNoBiasConfig {
-            ws_init: Init::KaimingUniform,
+            ws_init: DEFAULT_KAIMING_UNIFORM,
         }
     }
 }
diff --git a/src/fnet/fnet_model.rs b/src/fnet/fnet_model.rs
index 063c3d6..46c137b 100644
--- a/src/fnet/fnet_model.rs
+++ b/src/fnet/fnet_model.rs
@@ -1070,6 +1070,6 @@ mod test {
         let vs = tch::nn::VarStore::new(device);
         let config = FNetConfig::from_file(config_path);
 
-        let _: Box<dyn Send> = Box::new(FNetModel::new(&vs.root(), &config, true));
+        let _: Box<dyn Send> = Box::new(FNetModel::new(vs.root(), &config, true));
     }
 }
diff --git a/src/gpt2/gpt2_model.rs b/src/gpt2/gpt2_model.rs
index cebdf09..cbf442d 100644
--- a/src/gpt2/gpt2_model.rs
+++ b/src/gpt2/gpt2_model.rs
@@ -742,7 +742,7 @@ impl GPT2Generator {
         let mut var_store = nn::VarStore::new(device);
 
         let config = Gpt2Config::from_file(config_path);
-        let model = GPT2LMHeadModel::new(&var_store.root(), &config);
+        let model = GPT2LMHeadModel::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = tokenizer.get_bos_id();
diff --git a/src/gpt_neo/gpt_neo_model.rs b/src/gpt_neo/gpt_neo_model.rs
index 795e3bb..97f50fe 100644
--- a/src/gpt_neo/gpt_neo_model.rs
+++ b/src/gpt_neo/gpt_neo_model.rs
@@ -716,7 +716,7 @@ impl GptNeoGenerator {
         generate_config.validate();
         let mut var_store = nn::VarStore::new(device);
         let config = GptNeoConfig::from_file(config_path);
-        let model = GptNeoForCausalLM::new(&var_store.root(), &config)?;
+        let model = GptNeoForCausalLM::new(var_store.root(), &config)?;
         var_store.load(weights_path)?;
 
         let bos_token_id = tokenizer.get_bos_id();
diff --git a/src/m2m_100/embeddings.rs b/src/m2m_100/embeddings.rs
index 8531999..be46de6 100644
--- a/src/m2m_100/embeddings.rs
+++ b/src/m2m_100/embeddings.rs
@@ -71,7 +71,7 @@ impl SinusoidalPositionalEmbedding {
     ) -> Tensor {
         let half_dim = embedding_dim / 2;
 
-        let emb = -(10000f64.ln() as f64) / ((half_dim - 1) as f64);
+        let emb = -(10000f64.ln()) / ((half_dim - 1) as f64);
         let emb = (Tensor::arange(half_dim, (Kind::Float, device)) * emb).exp();
         let emb =
             Tensor::arange(num_embeddings, (Kind::Float, device)).unsqueeze(1) * emb.unsqueeze(0);
diff --git a/src/m2m_100/m2m_100_model.rs b/src/m2m_100/m2m_100_model.rs
index f9d8e24..0bffbe8 100644
--- a/src/m2m_100/m2m_100_model.rs
+++ b/src/m2m_100/m2m_100_model.rs
@@ -651,7 +651,7 @@ impl M2M100Generator {
         let mut var_store = nn::VarStore::new(device);
 
         let config = M2M100Config::from_file(config_path);
-        let model = M2M100ForConditionalGeneration::new(&var_store.root(), &config);
+        let model = M2M100ForConditionalGeneration::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = Some(config.bos_token_id.unwrap_or(0));
@@ -681,7 +681,7 @@ impl M2M100Generator {
     }
 
     fn force_token_id_generation(&self, scores: &mut Tensor, token_ids: &[i64]) {
-        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size() as i64)
+        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size())
             .filter(|pos| !token_ids.contains(pos))
             .collect();
         let impossible_tokens = Tensor::of_slice(&impossible_tokens).to_device(scores.device());
@@ -887,6 +887,6 @@ mod test {
         let vs = tch::nn::VarStore::new(device);
         let config = M2M100Config::from_file(config_path);
 
-        let _: Box<dyn Send> = Box::new(M2M100Model::new(&vs.root(), &config));
+        let _: Box<dyn Send> = Box::new(M2M100Model::new(vs.root(), &config));
     }
 }
diff --git a/src/marian/marian_model.rs b/src/marian/marian_model.rs
index 6368d01..a2ea85c 100644
--- a/src/marian/marian_model.rs
+++ b/src/marian/marian_model.rs
@@ -872,7 +872,7 @@ impl MarianGenerator {
         let mut var_store = nn::VarStore::new(device);
 
         let config = BartConfig::from_file(config_path);
-        let model = MarianForConditionalGeneration::new(&var_store.root(), &config);
+        let model = MarianForConditionalGeneration::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = Some(config.bos_token_id.unwrap_or(0));
@@ -904,7 +904,7 @@ impl MarianGenerator {
     }
 
     fn force_token_id_generation(&self, scores: &mut Tensor, token_ids: &[i64]) {
-        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size() as i64)
+        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size())
             .filter(|pos| !token_ids.contains(pos))
             .collect();
         let impossible_tokens = Tensor::of_slice(&impossible_tokens).to_device(scores.device());
diff --git a/src/mbart/mbart_model.rs b/src/mbart/mbart_model.rs
index 5a7cc72..f09c706 100644
--- a/src/mbart/mbart_model.rs
+++ b/src/mbart/mbart_model.rs
@@ -900,7 +900,7 @@ impl MBartGenerator {
         let mut var_store = nn::VarStore::new(device);
 
         let config = MBartConfig::from_file(config_path);
-        let model = MBartForConditionalGeneration::new(&var_store.root(), &config);
+        let model = MBartForConditionalGeneration::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = Some(config.bos_token_id.unwrap_or(0));
@@ -930,7 +930,7 @@ impl MBartGenerator {
     }
 
     fn force_token_id_generation(&self, scores: &mut Tensor, token_ids: &[i64]) {
-        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size() as i64)
+        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size())
             .filter(|pos| !token_ids.contains(pos))
             .collect();
         let impossible_tokens = Tensor::of_slice(&impossible_tokens).to_device(scores.device());
@@ -1136,6 +1136,6 @@ mod test {
         let vs = tch::nn::VarStore::new(device);
         let config = MBartConfig::from_file(config_path);
 
-        let _: Box<dyn Send> = Box::new(MBartModel::new(&vs.root(), &config));
+        let _: Box<dyn Send> = Box::new(MBartModel::new(vs.root(), &config));
     }
 }
diff --git a/src/mobilebert/mobilebert_model.rs b/src/mobilebert/mobilebert_model.rs
index 23ea8d7..5cd836c 100644
--- a/src/mobilebert/mobilebert_model.rs
+++ b/src/mobilebert/mobilebert_model.rs
@@ -19,6 +19,7 @@ use crate::{Config, RustBertError};
 use serde::{Deserialize, Serialize};
 use std::borrow::Borrow;
 use std::collections::HashMap;
+use tch::nn::init::DEFAULT_KAIMING_UNIFORM;
 use tch::nn::{Init, LayerNormConfig, Module};
 use tch::{nn, Kind, Tensor};
 
@@ -292,7 +293,7 @@ impl MobileBertLMPredictionHead {
                 config.hidden_size - config.embedding_size,
                 config.vocab_size,
             ],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
         let bias = p.var("bias", &[config.vocab_size], Init::Const(0.0));
 
diff --git a/src/openai_gpt/openai_gpt_model.rs b/src/openai_gpt/openai_gpt_model.rs
index 038809c..52adde7 100644
--- a/src/openai_gpt/openai_gpt_model.rs
+++ b/src/openai_gpt/openai_gpt_model.rs
@@ -504,7 +504,7 @@ impl OpenAIGenerator {
 
         let mut var_store = nn::VarStore::new(device);
         let config = Gpt2Config::from_file(config_path);
-        let model = OpenAIGPTLMHeadModel::new(&var_store.root(), &config);
+        let model = OpenAIGPTLMHeadModel::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = tokenizer.get_bos_id();
diff --git a/src/pegasus/pegasus_model.rs b/src/pegasus/pegasus_model.rs
index df1bbba..bed3108 100644
--- a/src/pegasus/pegasus_model.rs
+++ b/src/pegasus/pegasus_model.rs
@@ -624,7 +624,7 @@ impl PegasusConditionalGenerator {
         generate_config.validate();
         let mut var_store = nn::VarStore::new(device);
         let config = PegasusConfig::from_file(config_path);
-        let model = PegasusForConditionalGeneration::new(&var_store.root(), &config);
+        let model = PegasusForConditionalGeneration::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = Some(config.bos_token_id.unwrap_or(0));
@@ -654,7 +654,7 @@ impl PegasusConditionalGenerator {
     }
 
     fn force_token_id_generation(&self, scores: &mut Tensor, token_ids: &[i64]) {
-        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size() as i64)
+        let impossible_tokens: Vec<i64> = (0..self.get_vocab_size())
             .filter(|pos| !token_ids.contains(pos))
             .collect();
         let impossible_tokens = Tensor::of_slice(&impossible_tokens).to_device(scores.device());
diff --git a/src/pipelines/generation_utils.rs b/src/pipelines/generation_utils.rs
index d9a4ee3..710365c 100644
--- a/src/pipelines/generation_utils.rs
+++ b/src/pipelines/generation_utils.rs
@@ -403,7 +403,7 @@ pub(crate) mod private_generation_utils {
             prev_output_tokens: &Tensor,
             repetition_penalty: f64,
         ) {
-            for i in 0..(batch_size * num_beams as i64) {
+            for i in 0..(batch_size * num_beams) {
                 for token_position in 0..prev_output_tokens.get(i).size()[0] {
                     let token = prev_output_tokens.get(i).int64_value(&[token_position]);
                     let updated_value = &next_token_logits.double_value(&[i, token]);
@@ -826,8 +826,8 @@ pub(crate) mod private_generation_utils {
                 if gen_opt.no_repeat_ngram_size > 0 {
                     let banned_tokens = self.get_banned_tokens(
                         &input_ids,
-                        gen_opt.no_repeat_ngram_size as i64,
-                        current_length as i64,
+                        gen_opt.no_repeat_ngram_size,
+                        current_length,
                     );
                     for (batch_index, index_banned_token) in
                         (0..banned_tokens.len() as i64).zip(banned_tokens)
@@ -875,7 +875,7 @@ pub(crate) mod private_generation_utils {
                     }
                     self.top_k_top_p_filtering(
                         &mut next_token_logits,
-                        gen_opt.top_k as i64,
+                        gen_opt.top_k,
                         gen_opt.top_p,
                         1,
                     );
@@ -915,7 +915,7 @@ pub(crate) mod private_generation_utils {
                             &sentence_with_eos
                                 .to_kind(Kind::Bool)
                                 .to_device(sentence_lengths.device()),
-                            current_length as i64 + 1,
+                            current_length + 1,
                         );
                         unfinished_sentences = -unfinished_sentences * (sentence_with_eos - 1);
                     }
@@ -943,7 +943,7 @@ pub(crate) mod private_generation_utils {
                             &unfinished_sentences
                                 .to_kind(Kind::Bool)
                                 .to_device(sentence_lengths.device()),
-                            current_length as i64,
+                            current_length,
                         );
                         break;
                     }
@@ -1927,10 +1927,7 @@ pub trait LanguageGenerator<T: LMHeadModel, V: Vocab, U: Tokenizer<V>>:
         let batch_size = *input_ids.size().first().unwrap();
 
         let (effective_batch_size, effective_batch_mult) = match do_sample {
-            true => (
-                batch_size * num_return_sequences as i64,
-                num_return_sequences as i64,
-            ),
+            true => (batch_size * num_return_sequences, num_return_sequences),
             false => (batch_size, 1),
         };
 
@@ -1946,7 +1943,7 @@ pub trait LanguageGenerator<T: LMHeadModel, V: Vocab, U: Tokenizer<V>>:
             let encoder_outputs = self.encode(&input_ids, Some(&attention_mask)).unwrap();
             let expanded_batch_indices = Tensor::arange(batch_size, (Int64, input_ids.device()))
                 .view((-1, 1))
-                .repeat(&[1, num_beams as i64 * effective_batch_mult])
+                .repeat(&[1, num_beams * effective_batch_mult])
                 .view(-1);
             Some(encoder_outputs.index_select(0, &expanded_batch_indices))
         } else {
@@ -1959,19 +1956,19 @@ pub trait LanguageGenerator<T: LMHeadModel, V: Vocab, U: Tokenizer<V>>:
                     input_ids
                         .unsqueeze(1)
                         .expand(
-                            &[batch_size, effective_batch_mult * num_beams as i64, cur_len],
+                            &[batch_size, effective_batch_mult * num_beams, cur_len],
                             true,
                         )
                         .contiguous()
-                        .view((effective_batch_size * num_beams as i64, cur_len)),
+                        .view((effective_batch_size * num_beams, cur_len)),
                     attention_mask
                         .unsqueeze(1)
                         .expand(
-                            &[batch_size, effective_batch_mult * num_beams as i64, cur_len],
+                            &[batch_size, effective_batch_mult * num_beams, cur_len],
                             true,
                         )
                         .contiguous()
-                        .view((effective_batch_size * num_beams as i64, cur_len)),
+                        .view((effective_batch_size * num_beams, cur_len)),
                 )
             } else {
                 (input_ids, attention_mask)
@@ -1982,7 +1979,7 @@ pub trait LanguageGenerator<T: LMHeadModel, V: Vocab, U: Tokenizer<V>>:
                     .expect("decoder start id must be specified for encoder decoders")
             });
             let input_ids = Tensor::full(
-                &[effective_batch_size * num_beams as i64, 1],
+                &[effective_batch_size * num_beams, 1],
                 decoder_start_token_id,
                 (Int64, input_ids.device()),
             );
@@ -1990,15 +1987,11 @@ pub trait LanguageGenerator<T: LMHeadModel, V: Vocab, U: Tokenizer<V>>:
                 attention_mask
                     .unsqueeze(1)
                     .expand(
-                        &[
-                            batch_size,
-                            effective_batch_mult * num_beams as i64,
-                            input_ids_len,
-                        ],
+                        &[batch_size, effective_batch_mult * num_beams, input_ids_len],
                         true,
                     )
                     .contiguous()
-                    .view((effective_batch_size * num_beams as i64, input_ids_len))
+                    .view((effective_batch_size * num_beams, input_ids_len))
             } else {
                 attention_mask
             };
diff --git a/src/pipelines/keywords_extraction/pipeline.rs b/src/pipelines/keywords_extraction/pipeline.rs
index 51ccd86..3f74825 100644
--- a/src/pipelines/keywords_extraction/pipeline.rs
+++ b/src/pipelines/keywords_extraction/pipeline.rs
@@ -144,13 +144,13 @@ impl<'a> KeywordExtractionModel<'a> {
         config: KeywordExtractionConfig<'a>,
     ) -> Result<KeywordExtractionModel<'a>, RustBertError> {
         let tokenizer_config = SentenceEmbeddingsTokenizerConfig::from_file(
-            &config
+            config
                 .sentence_embeddings_config
                 .tokenizer_config_resource
                 .get_local_path()?,
         );
         let sentence_bert_config = SentenceEmbeddingsSentenceBertConfig::from_file(
-            &config
+            config
                 .sentence_embeddings_config
                 .sentence_bert_config_resource
                 .get_local_path()?,
diff --git a/src/pipelines/masked_language.rs b/src/pipelines/masked_language.rs
index 2ffad11..600446d 100644
--- a/src/pipelines/masked_language.rs
+++ b/src/pipelines/masked_language.rs
@@ -407,7 +407,7 @@ impl MaskedLanguageModel {
             .unwrap_or(usize::MAX);
 
         let language_encode =
-            MaskedLanguageOption::new(config.model_type, &var_store.root(), &model_config)?;
+            MaskedLanguageOption::new(config.model_type, var_store.root(), &model_config)?;
         var_store.load(weights_path)?;
         let mask_token = config.mask_token;
         Ok(MaskedLanguageModel {
diff --git a/src/pipelines/question_answering.rs b/src/pipelines/question_answering.rs
index 742501e..d6b6c30 100644
--- a/src/pipelines/question_answering.rs
+++ b/src/pipelines/question_answering.rs
@@ -620,7 +620,7 @@ impl QuestionAnsweringModel {
 
         let qa_model = QuestionAnsweringOption::new(
             question_answering_config.model_type,
-            &var_store.root(),
+            var_store.root(),
             &model_config,
         )?;
 
@@ -878,7 +878,7 @@ impl QuestionAnsweringModel {
             max_seq_length - sequence_pair_added_tokens - encoded_query.ids.len();
 
         let mut start_token = 0_usize;
-        while (spans.len() * doc_stride as usize) < encoded_context.ids.len() {
+        while (spans.len() * doc_stride) < encoded_context.ids.len() {
             let end_token = min(start_token + max_context_length, encoded_context.ids.len());
             let sub_encoded_context = TokenIdsWithOffsets {
                 ids: encoded_context.ids[start_token..end_token].to_vec(),
diff --git a/src/pipelines/sentence_embeddings/layers.rs b/src/pipelines/sentence_embeddings/layers.rs
index fffdc7a..fa63282 100644
--- a/src/pipelines/sentence_embeddings/layers.rs
+++ b/src/pipelines/sentence_embeddings/layers.rs
@@ -130,7 +130,7 @@ impl Dense {
             bias: dense_conf.bias,
         };
         let linear = nn::linear(
-            &vs_dense.root(),
+            vs_dense.root(),
             dense_conf.in_features,
             dense_conf.out_features,
             linear_conf,
diff --git a/src/pipelines/sentence_embeddings/pipeline.rs b/src/pipelines/sentence_embeddings/pipeline.rs
index 29e2e52..fa4d35b 100644
--- a/src/pipelines/sentence_embeddings/pipeline.rs
+++ b/src/pipelines/sentence_embeddings/pipeline.rs
@@ -230,11 +230,8 @@ impl SentenceEmbeddingsModel {
             transformer_type,
             transformer_config_resource.get_local_path()?,
         );
-        let transformer = SentenceEmbeddingsOption::new(
-            transformer_type,
-            &var_store.root(),
-            &transformer_config,
-        )?;
+        let transformer =
+            SentenceEmbeddingsOption::new(transformer_type, var_store.root(), &transformer_config)?;
         var_store.load(transformer_weights_resource.get_local_path()?)?;
 
         // Setup pooling layer
@@ -310,7 +307,7 @@ impl SentenceEmbeddingsModel {
                 Tensor::of_slice(
                     &input
                         .iter()
-                        .map(|&e| if e == pad_token_id { 0_i64 } else { 1_i64 })
+                        .map(|&e| i64::from(e != pad_token_id))
                         .collect::<Vec<_>>(),
                 )
             })
diff --git a/src/pipelines/sequence_classification.rs b/src/pipelines/sequence_classification.rs
index 700fb27..db241f3 100644
--- a/src/pipelines/sequence_classification.rs
+++ b/src/pipelines/sequence_classification.rs
@@ -593,7 +593,7 @@ impl SequenceClassificationModel {
             .map(|v| v as usize)
             .unwrap_or(usize::MAX);
         let sequence_classifier =
-            SequenceClassificationOption::new(config.model_type, &var_store.root(), &model_config)?;
+            SequenceClassificationOption::new(config.model_type, var_store.root(), &model_config)?;
         let label_mapping = model_config.get_label_mapping().clone();
         var_store.load(weights_path)?;
         Ok(SequenceClassificationModel {
diff --git a/src/pipelines/token_classification.rs b/src/pipelines/token_classification.rs
index 59cd1e3..7766ceb 100644
--- a/src/pipelines/token_classification.rs
+++ b/src/pipelines/token_classification.rs
@@ -699,7 +699,7 @@ impl TokenClassificationModel {
             .map(|v| v as usize)
             .unwrap_or(usize::MAX);
         let token_sequence_classifier =
-            TokenClassificationOption::new(config.model_type, &var_store.root(), &model_config)?;
+            TokenClassificationOption::new(config.model_type, var_store.root(), &model_config)?;
         let label_mapping = model_config.get_label_mapping().clone();
         let batch_size = config.batch_size;
         var_store.load(weights_path)?;
@@ -749,7 +749,7 @@ impl TokenClassificationModel {
         let mut start_token = 0_usize;
         let total_length = encoded_input.ids.len();
 
-        while (spans.len() * doc_stride as usize) < encoded_input.ids.len() {
+        while (spans.len() * doc_stride) < encoded_input.ids.len() {
             let end_token = min(start_token + max_content_length, total_length);
             let sub_encoded_input = TokenIdsWithOffsets {
                 ids: encoded_input.ids[start_token..end_token].to_vec(),
@@ -994,8 +994,8 @@ impl TokenClassificationModel {
         position_idx: i64,
         word_index: u16,
     ) -> Token {
-        let label_id = labels.int64_value(&[position_idx as i64]);
-        let token_id = input_tensor.int64_value(&[sentence_idx, position_idx as i64]);
+        let label_id = labels.int64_value(&[position_idx]);
+        let token_id = input_tensor.int64_value(&[sentence_idx, position_idx]);
 
         let offsets = &sentence_tokens.offsets[position_idx as usize];
 
diff --git a/src/pipelines/zero_shot_classification.rs b/src/pipelines/zero_shot_classification.rs
index c538324..0a69f40 100644
--- a/src/pipelines/zero_shot_classification.rs
+++ b/src/pipelines/zero_shot_classification.rs
@@ -575,7 +575,7 @@ impl ZeroShotClassificationModel {
         let mut var_store = VarStore::new(device);
         let model_config = ConfigOption::from_file(config.model_type, config_path);
         let zero_shot_classifier =
-            ZeroShotClassificationOption::new(config.model_type, &var_store.root(), &model_config)?;
+            ZeroShotClassificationOption::new(config.model_type, var_store.root(), &model_config)?;
         var_store.load(weights_path)?;
         Ok(ZeroShotClassificationModel {
             tokenizer,
diff --git a/src/prophetnet/decoder.rs b/src/prophetnet/decoder.rs
index 434d4a0..6e83c12 100644
--- a/src/prophetnet/decoder.rs
+++ b/src/prophetnet/decoder.rs
@@ -21,7 +21,7 @@ use crate::prophetnet::embeddings::ProphetNetPositionalEmbeddings;
 use crate::prophetnet::ProphetNetConfig;
 use crate::RustBertError;
 use std::borrow::{Borrow, BorrowMut};
-use tch::nn::Init;
+use tch::nn::init::DEFAULT_KAIMING_UNIFORM;
 use tch::{nn, Device, Kind, Tensor};
 
 fn ngram_attention_bias(sequence_length: i64, ngram: i64, device: Device, kind: Kind) -> Tensor {
@@ -210,7 +210,7 @@ impl ProphetNetDecoder {
         let ngram_embeddings = p_ngram_embedding.var(
             "weight",
             &[config.ngram, config.hidden_size],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
 
         let output_attentions = config.output_attentions.unwrap_or(false);
diff --git a/src/prophetnet/prophetnet_model.rs b/src/prophetnet/prophetnet_model.rs
index c035c64..cb3efd2 100644
--- a/src/prophetnet/prophetnet_model.rs
+++ b/src/prophetnet/prophetnet_model.rs
@@ -965,7 +965,7 @@ impl ProphetNetConditionalGenerator {
         generate_config.validate();
         let mut var_store = nn::VarStore::new(device);
         let config = ProphetNetConfig::from_file(config_path);
-        let model = ProphetNetForConditionalGeneration::new(&var_store.root(), &config)?;
+        let model = ProphetNetForConditionalGeneration::new(var_store.root(), &config)?;
         var_store.load(weights_path)?;
 
         let bos_token_id = Some(config.bos_token_id);
diff --git a/src/reformer/reformer_model.rs b/src/reformer/reformer_model.rs
index fb0b304..c367254 100644
--- a/src/reformer/reformer_model.rs
+++ b/src/reformer/reformer_model.rs
@@ -354,7 +354,7 @@ impl ReformerModel {
 
         let must_pad_to_match_chunk_length =
             (input_shape.last().unwrap() % self.least_common_mult_chunk_length != 0)
-                & (*input_shape.last().unwrap() as i64 > self.min_chunk_length)
+                & (*input_shape.last().unwrap() > self.min_chunk_length)
                 & old_layer_states.is_none();
 
         let start_idx_pos_encodings = if let Some(layer_states) = &old_layer_states {
@@ -1091,7 +1091,7 @@ impl ReformerGenerator {
         generate_config.validate();
         let mut var_store = nn::VarStore::new(device);
         let config = ReformerConfig::from_file(config_path);
-        let model = ReformerModelWithLMHead::new(&var_store.root(), &config)?;
+        let model = ReformerModelWithLMHead::new(var_store.root(), &config)?;
         var_store.load(weights_path)?;
 
         let bos_token_id = tokenizer.get_bos_id();
diff --git a/src/roberta/roberta_model.rs b/src/roberta/roberta_model.rs
index b384665..a96d72e 100644
--- a/src/roberta/roberta_model.rs
+++ b/src/roberta/roberta_model.rs
@@ -17,7 +17,7 @@ use crate::common::dropout::Dropout;
 use crate::common::linear::{linear_no_bias, LinearNoBias};
 use crate::roberta::embeddings::RobertaEmbeddings;
 use std::borrow::Borrow;
-use tch::nn::Init;
+use tch::nn::init::DEFAULT_KAIMING_UNIFORM;
 use tch::{nn, Tensor};
 
 /// # RoBERTa Pretrained model weight files
@@ -218,7 +218,7 @@ impl RobertaLMHead {
             config.vocab_size,
             Default::default(),
         );
-        let bias = p.var("bias", &[config.vocab_size], Init::KaimingUniform);
+        let bias = p.var("bias", &[config.vocab_size], DEFAULT_KAIMING_UNIFORM);
 
         RobertaLMHead {
             dense,
diff --git a/src/t5/t5_model.rs b/src/t5/t5_model.rs
index d33b3b9..3e40581 100644
--- a/src/t5/t5_model.rs
+++ b/src/t5/t5_model.rs
@@ -881,7 +881,7 @@ impl T5Generator {
         let mut var_store = nn::VarStore::new(device);
 
         let config = T5Config::from_file(config_path);
-        let model = T5ForConditionalGeneration::new(&var_store.root(), &config);
+        let model = T5ForConditionalGeneration::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = Some(config.bos_token_id.unwrap_or(-1));
diff --git a/src/xlnet/attention.rs b/src/xlnet/attention.rs
index f5e4cba..4492a58 100644
--- a/src/xlnet/attention.rs
+++ b/src/xlnet/attention.rs
@@ -15,7 +15,7 @@
 use crate::common::dropout::Dropout;
 use crate::xlnet::XLNetConfig;
 use std::borrow::Borrow;
-use tch::nn::Init;
+use tch::nn::init::DEFAULT_KAIMING_UNIFORM;
 use tch::{nn, Kind, Tensor};
 
 #[derive(Debug)]
@@ -72,52 +72,52 @@ impl XLNetRelativeAttention {
         let query = p.var(
             "q",
             &[config.d_model, config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
 
         let key = p.var(
             "k",
             &[config.d_model, config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
 
         let value = p.var(
             "v",
             &[config.d_model, config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
 
         let output = p.var(
             "o",
             &[config.d_model, config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
 
         let pos = p.var(
             "r",
             &[config.d_model, config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
 
         let r_r_bias = p.var(
             "r_r_bias",
             &[config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
         let r_s_bias = p.var(
             "r_s_bias",
             &[config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
         let r_w_bias = p.var(
             "r_w_bias",
             &[config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
         let seg_embed = p.var(
             "seg_embed",
             &[2, config.n_head, config.d_head],
-            Init::KaimingUniform,
+            DEFAULT_KAIMING_UNIFORM,
         );
 
         let dropout = Dropout::new(config.dropout);
diff --git a/src/xlnet/xlnet_model.rs b/src/xlnet/xlnet_model.rs
index 38ab301..60a0d7b 100644
--- a/src/xlnet/xlnet_model.rs
+++ b/src/xlnet/xlnet_model.rs
@@ -1648,7 +1648,7 @@ impl XLNetGenerator {
         let mut var_store = nn::VarStore::new(device);
 
         let config = XLNetConfig::from_file(config_path);
-        let model = XLNetLMHeadModel::new(&var_store.root(), &config);
+        let model = XLNetLMHeadModel::new(var_store.root(), &config);
         var_store.load(weights_path)?;
 
         let bos_token_id = Some(config.bos_token_id);
diff --git a/tests/albert.rs b/tests/albert.rs
index 9d787ec..b95b040 100644
--- a/tests/albert.rs
+++ b/tests/albert.rs
@@ -35,7 +35,7 @@ fn albert_masked_lm() -> anyhow::Result<()> {
     let tokenizer: AlbertTokenizer =
         AlbertTokenizer::from_file(vocab_path.to_str().unwrap(), true, false)?;
     let config = AlbertConfig::from_file(config_path);
-    let albert_model = AlbertForMaskedLM::new(&vs.root(), &config);
+    let albert_model = AlbertForMaskedLM::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -109,7 +109,7 @@ fn albert_for_sequence_classification() -> anyhow::Result<()> {
     config.id2label = Some(dummy_label_mapping);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let albert_model = AlbertForSequenceClassification::new(&vs.root(), &config);
+    let albert_model = AlbertForSequenceClassification::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -170,7 +170,7 @@ fn albert_for_multiple_choice() -> anyhow::Result<()> {
     let mut config = AlbertConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let albert_model = AlbertForMultipleChoice::new(&vs.root(), &config);
+    let albert_model = AlbertForMultipleChoice::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -242,7 +242,7 @@ fn albert_for_token_classification() -> anyhow::Result<()> {
     config.id2label = Some(dummy_label_mapping);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let albert_model = AlbertForTokenClassification::new(&vs.root(), &config);
+    let albert_model = AlbertForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -303,7 +303,7 @@ fn albert_for_question_answering() -> anyhow::Result<()> {
     let mut config = AlbertConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let albert_model = AlbertForQuestionAnswering::new(&vs.root(), &config);
+    let albert_model = AlbertForQuestionAnswering::new(vs.root(), &config);
 
     //    Define input
     let input = [
diff --git a/tests/bert.rs b/tests/bert.rs
index d790922..318d4bb 100644
--- a/tests/bert.rs
+++ b/tests/bert.rs
@@ -35,7 +35,7 @@ fn bert_masked_lm() -> anyhow::Result<()> {
     let tokenizer: BertTokenizer =
         BertTokenizer::from_file(vocab_path.to_str().unwrap(), true, true)?;
     let config = BertConfig::from_file(config_path);
-    let bert_model = BertForMaskedLM::new(&vs.root(), &config);
+    let bert_model = BertForMaskedLM::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -162,7 +162,7 @@ fn bert_for_sequence_classification() -> anyhow::Result<()> {
     config.id2label = Some(dummy_label_mapping);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let bert_model = BertForSequenceClassification::new(&vs.root(), &config);
+    let bert_model = BertForSequenceClassification::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -219,7 +219,7 @@ fn bert_for_multiple_choice() -> anyhow::Result<()> {
     let mut config = BertConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let bert_model = BertForMultipleChoice::new(&vs.root(), &config);
+    let bert_model = BertForMultipleChoice::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -283,7 +283,7 @@ fn bert_for_token_classification() -> anyhow::Result<()> {
     config.id2label = Some(dummy_label_mapping);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let bert_model = BertForTokenClassification::new(&vs.root(), &config);
+    let bert_model = BertForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -340,7 +340,7 @@ fn bert_for_question_answering() -> anyhow::Result<()> {
     let mut config = BertConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let bert_model = BertForQuestionAnswering::new(&vs.root(), &config);
+    let bert_model = BertForQuestionAnswering::new(vs.root(), &config);
 
     //    Define input
     let input = [
diff --git a/tests/deberta.rs b/tests/deberta.rs
index 9cd8537..f696c14 100644
--- a/tests/deberta.rs
+++ b/tests/deberta.rs
@@ -41,7 +41,7 @@ fn deberta_natural_language_inference() -> anyhow::Result<()> {
         false,
     )?;
     let config = DebertaConfig::from_file(config_path);
-    let model = DebertaForSequenceClassification::new(&vs.root(), &config);
+    let model = DebertaForSequenceClassification::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -96,7 +96,7 @@ fn deberta_masked_lm() -> anyhow::Result<()> {
     let mut config = DebertaConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let deberta_model = DebertaForMaskedLM::new(&vs.root(), &config);
+    let deberta_model = DebertaForMaskedLM::new(vs.root(), &config);
 
     //    Generate random input
     let input_tensor = Tensor::randint(42, &[32, 128], (Kind::Int64, device));
@@ -170,7 +170,7 @@ fn deberta_for_token_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(2, String::from("PER"));
     dummy_label_mapping.insert(3, String::from("ORG"));
     config.id2label = Some(dummy_label_mapping);
-    let model = DebertaForTokenClassification::new(&vs.root(), &config);
+    let model = DebertaForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let inputs = ["Where's Paris?", "In Kentucky, United States"];
@@ -225,7 +225,7 @@ fn deberta_for_question_answering() -> anyhow::Result<()> {
         false,
     )?;
     let config = DebertaConfig::from_file(config_path);
-    let model = DebertaForQuestionAnswering::new(&vs.root(), &config);
+    let model = DebertaForQuestionAnswering::new(vs.root(), &config);
 
     //    Define input
     let inputs = ["Where's Paris?", "Paris is in In Kentucky, United States"];
diff --git a/tests/deberta_v2.rs b/tests/deberta_v2.rs
index 12ceedf..0a46ee9 100644
--- a/tests/deberta_v2.rs
+++ b/tests/deberta_v2.rs
@@ -22,7 +22,7 @@ fn deberta_v2_masked_lm() -> anyhow::Result<()> {
     let mut config = DebertaV2Config::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let deberta_model = DebertaV2ForMaskedLM::new(&vs.root(), &config);
+    let deberta_model = DebertaV2ForMaskedLM::new(vs.root(), &config);
 
     //    Generate random input
     let input_tensor = Tensor::randint(42, &[32, 128], (Kind::Int64, device));
@@ -88,7 +88,7 @@ fn deberta_v2_for_sequence_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(1, String::from("Neutral"));
     dummy_label_mapping.insert(2, String::from("Negative"));
     config.id2label = Some(dummy_label_mapping);
-    let model = DebertaV2ForSequenceClassification::new(&vs.root(), &config);
+    let model = DebertaV2ForSequenceClassification::new(vs.root(), &config);
 
     //    Define input
     let inputs = ["Where's Paris?", "In Kentucky, United States"];
@@ -142,7 +142,7 @@ fn deberta_v2_for_token_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(2, String::from("PER"));
     dummy_label_mapping.insert(3, String::from("ORG"));
     config.id2label = Some(dummy_label_mapping);
-    let model = DebertaV2ForTokenClassification::new(&vs.root(), &config);
+    let model = DebertaV2ForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let inputs = ["Where's Paris?", "In Kentucky, United States"];
@@ -190,7 +190,7 @@ fn deberta_v2_for_question_answering() -> anyhow::Result<()> {
     let tokenizer =
         DeBERTaV2Tokenizer::from_file(vocab_path.to_str().unwrap(), false, false, false)?;
     let config = DebertaV2Config::from_file(config_path);
-    let model = DebertaV2ForQuestionAnswering::new(&vs.root(), &config);
+    let model = DebertaV2ForQuestionAnswering::new(vs.root(), &config);
 
     //    Define input
     let inputs = ["Where's Paris?", "Paris is in In Kentucky, United States"];
diff --git a/tests/distilbert.rs b/tests/distilbert.rs
index ae9c563..52d14bb 100644
--- a/tests/distilbert.rs
+++ b/tests/distilbert.rs
@@ -61,7 +61,7 @@ fn distilbert_masked_lm() -> anyhow::Result<()> {
     let tokenizer: BertTokenizer =
         BertTokenizer::from_file(vocab_path.to_str().unwrap(), true, true)?;
     let config = DistilBertConfig::from_file(config_path);
-    let distil_bert_model = DistilBertModelMaskedLM::new(&vs.root(), &config);
+    let distil_bert_model = DistilBertModelMaskedLM::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -140,7 +140,7 @@ fn distilbert_for_question_answering() -> anyhow::Result<()> {
     let mut config = DistilBertConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let distil_bert_model = DistilBertForQuestionAnswering::new(&vs.root(), &config);
+    let distil_bert_model = DistilBertForQuestionAnswering::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -211,7 +211,7 @@ fn distilbert_for_token_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(2, String::from("PER"));
     dummy_label_mapping.insert(3, String::from("ORG"));
     config.id2label = Some(dummy_label_mapping);
-    let distil_bert_model = DistilBertForTokenClassification::new(&vs.root(), &config);
+    let distil_bert_model = DistilBertForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let input = [
diff --git a/tests/distilgpt2.rs b/tests/distilgpt2.rs
index a20bf91..1ca7096 100644
--- a/tests/distilgpt2.rs
+++ b/tests/distilgpt2.rs
@@ -37,7 +37,7 @@ fn distilgpt2_lm_model() -> anyhow::Result<()> {
         false,
     )?;
     let config = Gpt2Config::from_file(config_path);
-    let gpt2_model = GPT2LMHeadModel::new(&vs.root(), &config);
+    let gpt2_model = GPT2LMHeadModel::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
diff --git a/tests/electra.rs b/tests/electra.rs
index 0f25b34..be50622 100644
--- a/tests/electra.rs
+++ b/tests/electra.rs
@@ -32,7 +32,7 @@ fn electra_masked_lm() -> anyhow::Result<()> {
     let mut config = ElectraConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let electra_model = ElectraForMaskedLM::new(&vs.root(), &config);
+    let electra_model = ElectraForMaskedLM::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -114,7 +114,7 @@ fn electra_discriminator() -> anyhow::Result<()> {
     let tokenizer: BertTokenizer =
         BertTokenizer::from_file(vocab_path.to_str().unwrap(), true, true)?;
     let config = ElectraConfig::from_file(config_path);
-    let electra_model = ElectraDiscriminator::new(&vs.root(), &config);
+    let electra_model = ElectraDiscriminator::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
diff --git a/tests/fnet.rs b/tests/fnet.rs
index 64d36b3..d1323a3 100644
--- a/tests/fnet.rs
+++ b/tests/fnet.rs
@@ -30,7 +30,7 @@ fn fnet_masked_lm() -> anyhow::Result<()> {
     let tokenizer: FNetTokenizer =
         FNetTokenizer::from_file(vocab_path.to_str().unwrap(), false, false)?;
     let config = FNetConfig::from_file(config_path);
-    let fnet_model = FNetForMaskedLM::new(&vs.root(), &config);
+    let fnet_model = FNetForMaskedLM::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -138,7 +138,7 @@ fn fnet_for_multiple_choice() -> anyhow::Result<()> {
     let mut config = FNetConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let fnet_model = FNetForMultipleChoice::new(&vs.root(), &config);
+    let fnet_model = FNetForMultipleChoice::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -201,7 +201,7 @@ fn fnet_for_token_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(3, String::from("ORG"));
     config.id2label = Some(dummy_label_mapping);
     config.output_hidden_states = Some(true);
-    let fnet_model = FNetForTokenClassification::new(&vs.root(), &config);
+    let fnet_model = FNetForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -256,7 +256,7 @@ fn fnet_for_question_answering() -> anyhow::Result<()> {
         FNetTokenizer::from_file(vocab_path.to_str().unwrap(), false, false)?;
     let mut config = FNetConfig::from_file(config_path);
     config.output_hidden_states = Some(true);
-    let fnet_model = FNetForQuestionAnswering::new(&vs.root(), &config);
+    let fnet_model = FNetForQuestionAnswering::new(vs.root(), &config);
 
     //    Define input
     let input = [
diff --git a/tests/gpt2.rs b/tests/gpt2.rs
index 07beb02..143ccd2 100644
--- a/tests/gpt2.rs
+++ b/tests/gpt2.rs
@@ -35,7 +35,7 @@ fn gpt2_lm_model() -> anyhow::Result<()> {
         false,
     )?;
     let config = Gpt2Config::from_file(config_path);
-    let gpt2_model = GPT2LMHeadModel::new(&vs.root(), &config);
+    let gpt2_model = GPT2LMHeadModel::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
diff --git a/tests/gpt_neo.rs b/tests/gpt_neo.rs
index 37a8fce..3618e00 100644
--- a/tests/gpt_neo.rs
+++ b/tests/gpt_neo.rs
@@ -40,7 +40,7 @@ fn gpt_neo_lm() -> anyhow::Result<()> {
     let mut config = GptNeoConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let gpt_neo_model = GptNeoForCausalLM::new(&vs.root(), &config)?;
+    let gpt_neo_model = GptNeoForCausalLM::new(vs.root(), &config)?;
     vs.load(weights_path)?;
 
     //    Define input
diff --git a/tests/longformer.rs b/tests/longformer.rs
index dc235a1..b7295c7 100644
--- a/tests/longformer.rs
+++ b/tests/longformer.rs
@@ -197,7 +197,7 @@ fn longformer_for_sequence_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(1, String::from("Negative"));
     dummy_label_mapping.insert(3, String::from("Neutral"));
     config.id2label = Some(dummy_label_mapping);
-    let model = LongformerForSequenceClassification::new(&vs.root(), &config);
+    let model = LongformerForSequenceClassification::new(vs.root(), &config);
 
     //    Define input
     let input = ["Very positive sentence", "Second sentence input"];
@@ -258,7 +258,7 @@ fn longformer_for_multiple_choice() -> anyhow::Result<()> {
         false,
     )?;
     let config = LongformerConfig::from_file(config_path);
-    let model = LongformerForMultipleChoice::new(&vs.root(), &config);
+    let model = LongformerForMultipleChoice::new(vs.root(), &config);
 
     //    Define input
     let prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced.";
@@ -337,7 +337,7 @@ fn longformer_for_token_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(2, String::from("PER"));
     dummy_label_mapping.insert(3, String::from("ORG"));
     config.id2label = Some(dummy_label_mapping);
-    let model = LongformerForTokenClassification::new(&vs.root(), &config);
+    let model = LongformerForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let inputs = ["Where's Paris?", "In Kentucky, United States"];
diff --git a/tests/mobilebert.rs b/tests/mobilebert.rs
index 50c3d04..09a4b86 100644
--- a/tests/mobilebert.rs
+++ b/tests/mobilebert.rs
@@ -35,7 +35,7 @@ fn mobilebert_masked_model() -> anyhow::Result<()> {
     let mut config = MobileBertConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let mobilebert_model = MobileBertForMaskedLM::new(&vs.root(), &config);
+    let mobilebert_model = MobileBertForMaskedLM::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -130,7 +130,7 @@ fn mobilebert_for_sequence_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(1, String::from("Negative"));
     dummy_label_mapping.insert(3, String::from("Neutral"));
     config.id2label = Some(dummy_label_mapping);
-    let model = MobileBertForSequenceClassification::new(&vs.root(), &config);
+    let model = MobileBertForSequenceClassification::new(vs.root(), &config);
 
     //    Define input
     let input = ["Very positive sentence", "Second sentence input"];
@@ -176,7 +176,7 @@ fn mobilebert_for_multiple_choice() -> anyhow::Result<()> {
     let vs = nn::VarStore::new(device);
     let tokenizer = BertTokenizer::from_file(vocab_path.to_str().unwrap(), true, true)?;
     let config = MobileBertConfig::from_file(config_path);
-    let model = MobileBertForMultipleChoice::new(&vs.root(), &config);
+    let model = MobileBertForMultipleChoice::new(vs.root(), &config);
 
     //    Define input
     let prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced.";
@@ -240,7 +240,7 @@ fn mobilebert_for_token_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(2, String::from("PER"));
     dummy_label_mapping.insert(3, String::from("ORG"));
     config.id2label = Some(dummy_label_mapping);
-    let model = MobileBertForTokenClassification::new(&vs.root(), &config);
+    let model = MobileBertForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let inputs = ["Where's Paris?", "In Kentucky, United States"];
@@ -287,7 +287,7 @@ fn mobilebert_for_question_answering() -> anyhow::Result<()> {
     let vs = nn::VarStore::new(device);
     let tokenizer = BertTokenizer::from_file(vocab_path.to_str().unwrap(), true, true)?;
     let config = MobileBertConfig::from_file(config_path);
-    let model = MobileBertForQuestionAnswering::new(&vs.root(), &config);
+    let model = MobileBertForQuestionAnswering::new(vs.root(), &config);
 
     //    Define input
     let inputs = ["Where's Paris?", "Paris is in In Kentucky, United States"];
diff --git a/tests/openai_gpt.rs b/tests/openai_gpt.rs
index 7680cca..0e12f5d 100644
--- a/tests/openai_gpt.rs
+++ b/tests/openai_gpt.rs
@@ -39,7 +39,7 @@ fn openai_gpt_lm_model() -> anyhow::Result<()> {
         true,
     )?;
     let config = OpenAiGptConfig::from_file(config_path);
-    let openai_gpt = OpenAIGPTLMHeadModel::new(&vs.root(), &config);
+    let openai_gpt = OpenAIGPTLMHeadModel::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
diff --git a/tests/reformer.rs b/tests/reformer.rs
index 5ac52a8..50b0e1d 100644
--- a/tests/reformer.rs
+++ b/tests/reformer.rs
@@ -98,7 +98,7 @@ fn reformer_for_sequence_classification() -> anyhow::Result<()> {
     config.id2label = Some(dummy_label_mapping);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let reformer_model = ReformerForSequenceClassification::new(&vs.root(), &config)?;
+    let reformer_model = ReformerForSequenceClassification::new(vs.root(), &config)?;
 
     //    Define input
     let input = [
@@ -159,7 +159,7 @@ fn reformer_for_question_answering() -> anyhow::Result<()> {
     let mut config = ReformerConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let reformer_model = ReformerForQuestionAnswering::new(&vs.root(), &config)?;
+    let reformer_model = ReformerForQuestionAnswering::new(vs.root(), &config)?;
 
     //    Define input
     let input = [
diff --git a/tests/roberta.rs b/tests/roberta.rs
index dbabeec..eb8976c 100644
--- a/tests/roberta.rs
+++ b/tests/roberta.rs
@@ -41,7 +41,7 @@ fn roberta_masked_lm() -> anyhow::Result<()> {
         false,
     )?;
     let config = RobertaConfig::from_file(config_path);
-    let roberta_model = RobertaForMaskedLM::new(&vs.root(), &config);
+    let roberta_model = RobertaForMaskedLM::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -136,7 +136,7 @@ fn roberta_for_sequence_classification() -> anyhow::Result<()> {
     config.id2label = Some(dummy_label_mapping);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let roberta_model = RobertaForSequenceClassification::new(&vs.root(), &config);
+    let roberta_model = RobertaForSequenceClassification::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -201,7 +201,7 @@ fn roberta_for_multiple_choice() -> anyhow::Result<()> {
     let mut config = RobertaConfig::from_file(config_path);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let roberta_model = RobertaForMultipleChoice::new(&vs.root(), &config);
+    let roberta_model = RobertaForMultipleChoice::new(vs.root(), &config);
 
     //    Define input
     let input = [
@@ -273,7 +273,7 @@ fn roberta_for_token_classification() -> anyhow::Result<()> {
     config.id2label = Some(dummy_label_mapping);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let roberta_model = RobertaForTokenClassification::new(&vs.root(), &config);
+    let roberta_model = RobertaForTokenClassification::new(vs.root(), &config);
 
     //    Define input
     let input = [
diff --git a/tests/xlnet.rs b/tests/xlnet.rs
index 2561c29..ba0d6b5 100644
--- a/tests/xlnet.rs
+++ b/tests/xlnet.rs
@@ -141,7 +141,7 @@ fn xlnet_lm_model() -> anyhow::Result<()> {
     let tokenizer: XLNetTokenizer =
         XLNetTokenizer::from_file(vocab_path.to_str().unwrap(), false, true)?;
     let config = XLNetConfig::from_file(config_path);
-    let xlnet_model = XLNetLMHeadModel::new(&vs.root(), &config);
+    let xlnet_model = XLNetLMHeadModel::new(vs.root(), &config);
     vs.load(weights_path)?;
 
     //    Define input
@@ -257,7 +257,7 @@ fn xlnet_for_sequence_classification() -> anyhow::Result<()> {
     config.id2label = Some(dummy_label_mapping);
     config.output_attentions = Some(true);
     config.output_hidden_states = Some(true);
-    let xlnet_model = XLNetForSequenceClassification::new(&vs.root(), &config)?;
+    let xlnet_model = XLNetForSequenceClassification::new(vs.root(), &config)?;
 
     //    Define input
     let input = ["Very positive sentence", "Second sentence input"];
@@ -322,7 +322,7 @@ fn xlnet_for_multiple_choice() -> anyhow::Result<()> {
     let vs = nn::VarStore::new(device);
     let tokenizer = XLNetTokenizer::from_file(vocab_path.to_str().unwrap(), true, true)?;
     let config = XLNetConfig::from_file(config_path);
-    let xlnet_model = XLNetForMultipleChoice::new(&vs.root(), &config)?;
+    let xlnet_model = XLNetForMultipleChoice::new(vs.root(), &config)?;
 
     //    Define input
     let prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced.";
@@ -396,7 +396,7 @@ fn xlnet_for_token_classification() -> anyhow::Result<()> {
     dummy_label_mapping.insert(2, String::from("PER"));
     dummy_label_mapping.insert(3, String::from("ORG"));
     config.id2label = Some(dummy_label_mapping);
-    let xlnet_model = XLNetForTokenClassification::new(&vs.root(), &config)?;
+    let xlnet_model = XLNetForTokenClassification::new(vs.root(), &config)?;
 
     //    Define input
     let inputs = ["Where's Paris?", "In Kentucky, United States"];
@@ -453,7 +453,7 @@ fn xlnet_for_question_answering() -> anyhow::Result<()> {
     let vs = nn::VarStore::new(device);
     let tokenizer = XLNetTokenizer::from_file(vocab_path.to_str().unwrap(), true, true)?;
     let config = XLNetConfig::from_file(config_path);
-    let xlnet_model = XLNetForQuestionAnswering::new(&vs.root(), &config)?;
+    let xlnet_model = XLNetForQuestionAnswering::new(vs.root(), &config)?;
 
     //    Define input
     let inputs = ["Where's Paris?", "Paris is in In Kentucky, United States"];