Dependencies update & Clippy fixes (#483)

* wip dep & clippy updates

* clippy (2)

* pin ndarray to 0.15

* fmt

* clippy warnings

* clippy (3)

Author: guillaume-be

Cargo.toml

@@ -79,30 +79,30 @@ rust_tokenizers = "8.1.1"
 tch = { version = "0.17.0", features = ["download-libtorch"] }
 serde_json = "1"
 serde = { version = "1", features = ["derive"] }
-ordered-float = "4.2.0"
+ordered-float = "5"
 uuid = { version = "1", features = ["v4"] }
-thiserror = "1"
+thiserror = "2"
 half = "2"
-regex = "1.10"
+regex = "1.11"
 
-cached-path = { version = "0.6", default-features = false, optional = true }
-dirs = { version = "5", optional = true }
+cached-path = { version = "0.8", default-features = false, optional = true }
+dirs = { version = "6", optional = true }
 lazy_static = { version = "1", optional = true }
 ort = { version = "1.16.3", optional = true, default-features = false, features = [
     "half",
 ] }
 ndarray = { version = "0.15", optional = true }
-tokenizers = { version = "0.20", optional = true, default-features = false, features = [
+tokenizers = { version = "0.21", optional = true, default-features = false, features = [
     "onig",
 ] }
 
 [dev-dependencies]
 anyhow = "1"
 csv = "1"
-criterion = "0.5"
-tokio = { version = "1.35", features = ["sync", "rt-multi-thread", "macros"] }
+criterion = "0.6"
+tokio = { version = "1.45", features = ["sync", "rt-multi-thread", "macros"] }
 tempfile = "3"
-itertools = "0.13.0"
+itertools = "0.14.0"
 tracing-subscriber = { version = "0.3", default-features = false, features = [
     "env-filter",
     "fmt",

benches/generation_benchmark.rs

@@ -1,13 +1,14 @@
 #[macro_use]
 extern crate criterion;
 
-use criterion::{black_box, Criterion};
+use criterion::Criterion;
 use rust_bert::gpt2::{
     Gpt2ConfigResources, Gpt2MergesResources, Gpt2ModelResources, Gpt2VocabResources,
 };
 use rust_bert::pipelines::common::{ModelResource, ModelType};
 use rust_bert::pipelines::text_generation::{TextGenerationConfig, TextGenerationModel};
 use rust_bert::resources::RemoteResource;
+use std::hint::black_box;
 use std::time::{Duration, Instant};
 use tch::Device;
 

benches/squad_benchmark.rs

@@ -1,14 +1,15 @@
 #[macro_use]
 extern crate criterion;
 
-use criterion::{black_box, Criterion};
+use criterion::Criterion;
 use rust_bert::bert::{BertConfigResources, BertModelResources, BertVocabResources};
 use rust_bert::pipelines::common::{ModelResource, ModelType};
 use rust_bert::pipelines::question_answering::{
     squad_processor, QaInput, QuestionAnsweringConfig, QuestionAnsweringModel,
 };
 use rust_bert::resources::RemoteResource;
 use std::env;
+use std::hint::black_box;
 use std::path::PathBuf;
 use std::time::{Duration, Instant};
 

benches/summarization_benchmark.rs

@@ -1,8 +1,9 @@
 #[macro_use]
 extern crate criterion;
 
-use criterion::{black_box, Criterion};
+use criterion::Criterion;
 use rust_bert::pipelines::summarization::{SummarizationConfig, SummarizationModel};
+use std::hint::black_box;
 use std::time::{Duration, Instant};
 use tch::Device;
 

benches/tensor_operations_benchmark.rs

@@ -1,7 +1,8 @@
 #[macro_use]
 extern crate criterion;
 
-use criterion::{black_box, Criterion};
+use criterion::Criterion;
+use std::hint::black_box;
 use std::time::{Duration, Instant};
 use tch::kind::Kind;
 use tch::{Device, Tensor};

benches/token_classification_benchmark.rs

@@ -1,7 +1,8 @@
-use criterion::{black_box, criterion_group, criterion_main, Criterion};
+use criterion::{criterion_group, criterion_main, Criterion};
 use rust_bert::pipelines::token_classification::{
     TokenClassificationConfig, TokenClassificationModel,
 };
+use std::hint::black_box;
 use tch::Device;
 
 fn create_model() -> TokenClassificationModel {

benches/translation_benchmark.rs

@@ -1,7 +1,8 @@
 #[macro_use]
 extern crate criterion;
 
-use criterion::{black_box, Criterion};
+use criterion::Criterion;
+use std::hint::black_box;
 // use rust_bert::pipelines::common::ModelType;
 // use rust_bert::pipelines::translation::TranslationOption::{Marian, T5};
 use rust_bert::pipelines::common::ModelType;

examples/generation_gptj.rs

@@ -23,15 +23,15 @@ use tch::Device;
 /// inputs = tokenizer(prompts, return_tensors="pt", padding=True, truncation=True).to(device)
 ///
 /// with torch.no_grad():
-///     gen_tokens = model.generate(
-///         **inputs,
-///         min_length=0,
-///         max_length=32,
-///         do_sample=False,
-///         early_stopping=True,
-///         num_beams=1,
-///         num_return_sequences=1
-///     )
+///  gen_tokens = model.generate(
+///      **inputs,
+///      min_length=0,
+///      max_length=32,
+///      do_sample=False,
+///      early_stopping=True,
+///      num_beams=1,
+///      num_return_sequences=1
+///  )
 ///
 /// gen_texts = tokenizer.batch_decode(gen_tokens, skip_special_tokens=True)
 /// ````

src/common/resources/local.rs

@@ -22,7 +22,7 @@ impl ResourceProvider for LocalResource {
     /// use rust_bert::resources::{LocalResource, ResourceProvider};
     /// use std::path::PathBuf;
     /// let config_resource = LocalResource {
-    ///     local_path: PathBuf::from("path/to/config.json"),
+    ///  local_path: PathBuf::from("path/to/config.json"),
     /// };
     /// let config_path = config_resource.get_local_path();
     /// ```
@@ -42,7 +42,7 @@ impl ResourceProvider for LocalResource {
     /// use rust_bert::resources::{LocalResource, ResourceProvider};
     /// use std::path::PathBuf;
     /// let config_resource = LocalResource {
-    ///     local_path: PathBuf::from("path/to/config.json"),
+    ///  local_path: PathBuf::from("path/to/config.json"),
     /// };
     /// let config_path = config_resource.get_resource();
     /// ```

src/common/resources/mod.rs

@@ -12,7 +12,7 @@
 //! - LocalResource: points to a local file
 //! - RemoteResource: points to a remote file via a URL
 //! - BufferResource: refers to a buffer that contains file contents for a resource (currently only
-//!                   usable for weights)
+//!   usable for weights)
 //!
 //! For `LocalResource` and `RemoteResource`, the local location of the file can be retrieved using
 //! `get_local_path`, allowing to reference the resource file location regardless if it is a remote
@@ -52,7 +52,7 @@ pub trait ResourceProvider: Debug + Send + Sync {
     /// use rust_bert::resources::{LocalResource, ResourceProvider};
     /// use std::path::PathBuf;
     /// let config_resource = LocalResource {
-    ///     local_path: PathBuf::from("path/to/config.json"),
+    ///  local_path: PathBuf::from("path/to/config.json"),
     /// };
     /// let config_path = config_resource.get_local_path();
     /// ```