Fix typos (#10483)

Author: co63oc

CONTRIBUTING.md

@@ -72,7 +72,7 @@ PaddleNLP 使用 [Git 分支模型](http://nvie.com/posts/a-successful-git-branc
    类似于以下内容：
 
    ```text
-    ➜  (my-virtual-env) git commit -m "commiting my cool feature"
+    ➜  (my-virtual-env) git commit -m "committing my cool feature"
     black....................................................................Passed
     isort....................................................................Passed
     flake8...................................................................Passed
@@ -91,7 +91,7 @@ PaddleNLP 使用 [Git 分支模型](http://nvie.com/posts/a-successful-git-branc
    但大多数时候事情并没有那么顺利。当您的代码或文档不符合标准时，`pre-commit` 检查将失败。
 
    ```text
-    ➜  (my-virtual-env) git commit -m "commiting my cool feature"
+    ➜  (my-virtual-env) git commit -m "committing my cool feature"
     black....................................................................Passed
     isort....................................................................Failed
     - hook id: isort

csrc/gpu/moe/fused_moe/cutlass_kernels/moe_gemm/fused_moe_gemm_kernels_template.h

@@ -725,7 +725,7 @@ void MoeGemmRunner<T, WeightType>::run_gemm<EpilogueTag>(
       gemmConfigManager.addBestConfig(gemmId, profile_total_rows, best_config);
       chosen_config = best_config;
     } else {
-      PADDLE_FATAL("[MoE Configure Search] find no one avaliable config.");
+      PADDLE_FATAL("[MoE Configure Search] find no one available config.");
     }
   }
   dispatch_to_arch<EpilogueTag>(A,

paddlenlp/__init__.py

@@ -35,7 +35,7 @@
 
     logger.warning(
         "Detected that datasets module was imported before paddlenlp. "
-        "This may cause PaddleNLP datasets to be unavalible in intranet. "
+        "This may cause PaddleNLP datasets to be unavailable in intranet. "
         "Please import paddlenlp before datasets module to avoid download issues"
     )
 import paddle

paddlenlp/dataaug/char.py

@@ -436,7 +436,7 @@ def _augment_single(self, seq_tokens, aug_indexes):
         return sentences
 
     def _generate_sequence(self, output_seq_tokens, aug_tokens, p):
-        """Genearte the sequences according to the mapping list"""
+        """Generate the sequences according to the mapping list"""
         for aug_token in aug_tokens:
             idx, token = aug_token
             if p == 0:

paddlenlp/dataaug/sentence.py

@@ -33,9 +33,9 @@
 class SentenceGenerate:
     """
     SentenceGenerate is a sentence-level data augmentation strategy
-    that generates simialr sentences according to the input sequence.
-    The strattegy first generates several sentences, and then chooses
-    the top n simialr sentences by the model.
+    that generates similar sentences according to the input sequence.
+    The strategy first generates several sentences, and then chooses
+    the top n similar sentences by the model.
 
     Args:
         model_name (str):
@@ -82,7 +82,7 @@ def augment(self, sequences):
 
     @paddle.no_grad()
     def _generate_similar_sentence(self, sequence, model, tokenizer):
-        """Generates generate_n similar sentences from the provided sequence, and chooose the best create_n similar sentences."""
+        """Generates generate_n similar sentences from the provided sequence, and choose the best create_n similar sentences."""
 
         # Generate generate_n similar sentences
         generated_sequences = [sequence]
@@ -134,7 +134,7 @@ class SentenceSummarize:
             token probabilities in the "sampling" strategy. Default to 1.0,
             which means no effect.
         use_fp16_decoding: (bool): Whether to use fp16 for decoding.
-            Only works when faster entry is avalible. Default to False.
+            Only works when faster entry is available. Default to False.
         kwargs (dict): Additional keyword arguments refer to ..taskflow.text_summarization.TextSummarization
     """
 
@@ -337,8 +337,8 @@ def _translate(self, model, tokenizer, sequences, lang):
 class SentenceBackTranslateAPI:
     """
     SentenceBackTranslateAPI is a sentence-level data augmentation strategy
-    that translates the input sequence into one langugage, and backtranslate
-    back into the sourche language by baidu translate api.
+    that translates the input sequence into one language, and back-translate
+    back into the source language by baidu translate api.
 
     Args:
         src_lang (str):

paddlenlp/dataaug/word.py

@@ -170,7 +170,7 @@ def _load_substitute_dict(self, source_type):
         return substitute_dict
 
     def _generate_sequence(self, output_seq_tokens, aug_tokens):
-        """Genearte the sequences according to the mapping list"""
+        """Generate the sequences according to the mapping list"""
         for aug_token in aug_tokens:
             idx, token = aug_token
             output_seq_tokens[int(idx)] = token
@@ -503,7 +503,7 @@ def _augment_single(self, seq_tokens, aug_indexes):
         return sentences
 
     def _generate_sequence(self, output_seq_tokens, aug_tokens, p):
-        """Genearte the sequences according to the mapping list"""
+        """Generate the sequences according to the mapping list"""
         for aug_token in aug_tokens:
             idx, token = aug_token
             if p == 0:

paddlenlp/experimental/autonlp/README_en.md

@@ -50,8 +50,8 @@ Args:
 - text_column (string, required): Name of the column that contains the input text.
 - label_column (string, required): Name of the column that contains the target variable to predict.
 - language (string, required): language of the text
-- metric_for_best_model (string, optional): the name of the metrc for selecting the best model.
-- greater_is_better (bool, optional): Whether better models should have a greater metric or not. Use in conjuction with `metric_for_best_model`.
+- metric_for_best_model (string, optional): the name of the metric for selecting the best model.
+- greater_is_better (bool, optional): Whether better models should have a greater metric or not. Use in conjunction with `metric_for_best_model`.
 - problem_type (str, optional): Select among ["multi_class", "multi_label"] based on the nature of your problem
 - output_dir (str, optional): Output directory for the experiments, defaults to "autpnlp_results"
 - verbosity: (int, optional): controls the verbosity of the run. Defaults to 1, which let the workers log to the driver.To reduce the amount of logs, use verbosity > 0 to set stop the workers from logging to the driver.

paddlenlp/experimental/autonlp/auto_trainer_base.py

@@ -45,7 +45,7 @@ class AutoTrainerBase(metaclass=ABCMeta):
         eval_dataset (Dataset, required): Evaluation dataset, must contains the 'text_column' and 'label_column' specified below
         language (string, required): language of the text
         metric_for_best_model (string, optional): the name of the metric for selecting the best model.
-        greater_is_better (bool, required): Whether better models should have a greater metric or not. Use in conjuction with `metric_for_best_model`.
+        greater_is_better (bool, required): Whether better models should have a greater metric or not. Use in conjunction with `metric_for_best_model`.
         output_dir (str, optional): Output directory for the experiments, defaults to "autpnlp_results"
         verbosity: (int, optional): controls the verbosity of the run. Defaults to 1, which let the workers log to the driver.To reduce the amount of logs,
                 use verbosity > 0 to set stop the workers from logging to the driver.

paddlenlp/experimental/autonlp/text_classification.py

@@ -59,8 +59,8 @@ class AutoTrainerForTextClassification(AutoTrainerBase):
         eval_dataset (Dataset, required): Evaluation dataset, must contains the 'text_column' and 'label_column' specified below
         text_column (string, required): Name of the column that contains the input text.
         label_column (string, required): Name of the column that contains the target variable to predict.
-        metric_for_best_model (string, optional): the name of the metrc for selecting the best model. Default to 'eval_accuracy'.
-        greater_is_better (bool, optional): Whether better models should have a greater metric or not. Use in conjuction with `metric_for_best_model`.
+        metric_for_best_model (string, optional): the name of the metric for selecting the best model. Default to 'eval_accuracy'.
+        greater_is_better (bool, optional): Whether better models should have a greater metric or not. Use in conjunction with `metric_for_best_model`.
         problem_type (str, optional): Select among ["multi_class", "multi_label"] based on the nature of your problem
         kwargs (dict, optional): Additional keyword arguments passed along to the specific task.
             language (string, required): language of the text.
@@ -639,7 +639,7 @@ def export(self, export_path: str, trial_id: Optional[str] = None, compress: boo
             f"Taskflow config saved to {export_path}. You can use the Taskflow config to create a Taskflow instance for inference"
         )
 
-        logger.info(f"Exported trial_id: {trial_id} to export_path: {export_path} sucessfully!")
+        logger.info(f"Exported trial_id: {trial_id} to export_path: {export_path} successfully!")
 
         if os.path.exists(self.training_path):
             logger.info("Removing training checkpoints to conserve disk space")

paddlenlp/experimental/faster_tokenizer.py

@@ -29,10 +29,10 @@
 def to_tensor(string_values, name="text"):
     """
     Create the tensor that the value holds the list of string.
-    NOTICE: The value will be holded in the cpu place.
+    NOTICE: The value will be held in the cpu place.
 
     Args:
-        string_values(list[string]): The value will be setted to the tensor.
+        string_values(list[string]): The value will be set to the tensor.
         name(string): The name of the tensor.
     """
     tensor = paddle.Tensor(core.VarDesc.VarType.STRING, [], name, core.VarDesc.VarType.STRINGS, False)
@@ -43,7 +43,7 @@ def to_tensor(string_values, name="text"):
 def to_vocab_buffer(vocab_dict, name):
     """
     Create the tensor that the value holds the map, the type of key is the string.
-    NOTICE: The value will be holded in the cpu place.
+    NOTICE: The value will be held in the cpu place.
 
     Args:
         vocab_dict(dict): The value will be setted to the tensor.