diff --git a/fastNLP/core/metrics/accuracy.py b/fastNLP/core/metrics/accuracy.py
index 47d5e114..fbd826bd 100644
--- a/fastNLP/core/metrics/accuracy.py
+++ b/fastNLP/core/metrics/accuracy.py
@@ -69,7 +69,7 @@ class Accuracy(Metric):
         elif pred.ndim == target.ndim + 1:
             pred = pred.argmax(axis=-1)
             if seq_len is None and target.ndim > 1:
-                logger.warn("You are not passing `seq_len` to exclude pad when calculate accuracy.")
+                logger.warning("You are not passing `seq_len` to exclude pad when calculate accuracy.")
 
         else:
             raise RuntimeError(f"when pred have size:{pred.shape}, target should have size: {pred.shape} or "
diff --git a/fastNLP/core/metrics/classify_f1_pre_rec_metric.py b/fastNLP/core/metrics/classify_f1_pre_rec_metric.py
index daf325c0..39565f40 100644
--- a/fastNLP/core/metrics/classify_f1_pre_rec_metric.py
+++ b/fastNLP/core/metrics/classify_f1_pre_rec_metric.py
@@ -156,7 +156,7 @@ class ClassifyFPreRecMetric(Metric):
         elif pred.ndim == target.ndim + 1:
             pred = pred.argmax(axis=-1)
             if seq_len is None and target.ndim > 1:
-                logger.warn("You are not passing `seq_len` to exclude pad when calculate accuracy.")
+                logger.warning("You are not passing `seq_len` to exclude pad when calculate accuracy.")
         else:
             raise RuntimeError(f"when pred have "
                                f"size:{pred.shape}, target should have size: {pred.shape} or "
diff --git a/fastNLP/core/metrics/span_f1_pre_rec_metric.py b/fastNLP/core/metrics/span_f1_pre_rec_metric.py
index 9a0b1d9d..07a6cd56 100644
--- a/fastNLP/core/metrics/span_f1_pre_rec_metric.py
+++ b/fastNLP/core/metrics/span_f1_pre_rec_metric.py
@@ -39,7 +39,7 @@ def _check_tag_vocab_and_encoding_type(tag_vocab: Union[Vocabulary, dict], encod
                             f"encoding_type."
         tags = tags.replace(tag, '')  # 删除该值
     if tags:  # 如果不为空，说明出现了未使用的tag
-        logger.warn(f"Tag:{tags} in encoding type:{encoding_type} is not presented in your Vocabulary. Check your "
+        logger.warning(f"Tag:{tags} in encoding type:{encoding_type} is not presented in your Vocabulary. Check your "
                       "encoding_type.")
 
 
diff --git a/fastNLP/core/utils/utils.py b/fastNLP/core/utils/utils.py
index 33a7ee7e..ec0c87b0 100644
--- a/fastNLP/core/utils/utils.py
+++ b/fastNLP/core/utils/utils.py
@@ -554,7 +554,7 @@ def deprecated(help_message: Optional[str] = None):
         def wrapper(*args, **kwargs):
             func_hash = hash(deprecated_function)
             if func_hash not in _emitted_deprecation_warnings:
-                logger.warn(warning_msg, category=FutureWarning, stacklevel=2)
+                logger.warning(warning_msg, category=FutureWarning, stacklevel=2)
                 _emitted_deprecation_warnings.add(func_hash)
             return deprecated_function(*args, **kwargs)
 
diff --git a/fastNLP/embeddings/torch/static_embedding.py b/fastNLP/embeddings/torch/static_embedding.py
index de2b231a..cc15c214 100644
--- a/fastNLP/embeddings/torch/static_embedding.py
+++ b/fastNLP/embeddings/torch/static_embedding.py
@@ -286,7 +286,7 @@ class StaticEmbedding(TokenEmbedding):
                     if word in vocab:
                         index = vocab.to_index(word)
                         if index in matrix:
-                            logger.warn(f"Word has more than one vector in embedding file. Set logger level to "
+                            logger.warning(f"Word has more than one vector in embedding file. Set logger level to "
                                           f"DEBUG for detail.")
                             logger.debug(f"Word:{word} occurs again in line:{idx}(starts from 0)")
                         matrix[index] = torch.from_numpy(np.fromstring(' '.join(nums), sep=' ', dtype=dtype, count=dim))
@@ -295,7 +295,7 @@ class StaticEmbedding(TokenEmbedding):
                         found_count += 1
                 except Exception as e:
                     if error == 'ignore':
-                        logger.warn("Error occurred at the {} line.".format(idx))
+                        logger.warning("Error occurred at the {} line.".format(idx))
                     else:
                         logger.error("Error occurred at the {} line.".format(idx))
                         raise e
diff --git a/fastNLP/io/embed_loader.py b/fastNLP/io/embed_loader.py
index 9080ff28..df82643b 100644
--- a/fastNLP/io/embed_loader.py
+++ b/fastNLP/io/embed_loader.py
@@ -91,7 +91,7 @@ class EmbedLoader:
                         hit_flags[index] = True
                 except Exception as e:
                     if error == 'ignore':
-                        logger.warn("Error occurred at the {} line.".format(idx))
+                        logger.warning("Error occurred at the {} line.".format(idx))
                     else:
                         logging.error("Error occurred at the {} line.".format(idx))
                         raise e
@@ -156,7 +156,7 @@ class EmbedLoader:
                         found_pad = True
                 except Exception as e:
                     if error == 'ignore':
-                        logger.warn("Error occurred at the {} line.".format(idx))
+                        logger.warning("Error occurred at the {} line.".format(idx))
                         pass
                     else:
                         logging.error("Error occurred at the {} line.".format(idx))
diff --git a/fastNLP/io/loader/classification.py b/fastNLP/io/loader/classification.py
index 4416376f..2ae0b163 100644
--- a/fastNLP/io/loader/classification.py
+++ b/fastNLP/io/loader/classification.py
@@ -345,7 +345,7 @@ class SST2Loader(Loader):
         with open(path, 'r', encoding='utf-8') as f:
             f.readline()  # 跳过header
             if 'test' in os.path.split(path)[1]:
-                logger.warn("SST2's test file has no target.")
+                logger.warning("SST2's test file has no target.")
                 for line in f:
                     line = line.strip()
                     if line:
diff --git a/fastNLP/io/loader/matching.py b/fastNLP/io/loader/matching.py
index 5595b798..08387df9 100644
--- a/fastNLP/io/loader/matching.py
+++ b/fastNLP/io/loader/matching.py
@@ -55,7 +55,7 @@ class MNLILoader(Loader):
         with open(path, 'r', encoding='utf-8') as f:
             f.readline()  # 跳过header
             if path.endswith("test_matched.tsv") or path.endswith('test_mismatched.tsv'):
-                logger.warn("MNLI's test file has no target.")
+                logger.warning("MNLI's test file has no target.")
                 for line in f:
                     line = line.strip()
                     if line:
@@ -227,7 +227,7 @@ class QNLILoader(JsonLoader):
         with open(path, 'r', encoding='utf-8') as f:
             f.readline()  # 跳过header
             if path.endswith("test.tsv"):
-                logger.warn("QNLI's test file has no target.")
+                logger.warning("QNLI's test file has no target.")
                 for line in f:
                     line = line.strip()
                     if line:
@@ -289,7 +289,7 @@ class RTELoader(Loader):
         with open(path, 'r', encoding='utf-8') as f:
             f.readline()  # 跳过header
             if path.endswith("test.tsv"):
-                logger.warn("RTE's test file has no target.")
+                logger.warning("RTE's test file has no target.")
                 for line in f:
                     line = line.strip()
                     if line:
diff --git a/fastNLP/io/pipe/matching.py b/fastNLP/io/pipe/matching.py
index a89f2f2b..baebdbaa 100644
--- a/fastNLP/io/pipe/matching.py
+++ b/fastNLP/io/pipe/matching.py
@@ -146,7 +146,7 @@ class MatchingBertPipe(Pipe):
             warn_msg = f"There are {len(target_vocab._no_create_word)} target labels" \
                        f" in {[name for name in data_bundle.datasets.keys() if 'train' not in name]} " \
                        f"data set but not in train data set!."
-            logger.warn(warn_msg)
+            logger.warning(warn_msg)
             print(warn_msg)
         
         has_target_datasets = [dataset for name, dataset in data_bundle.datasets.items() if
@@ -291,7 +291,7 @@ class MatchingPipe(Pipe):
             warn_msg = f"There are {len(target_vocab._no_create_word)} target labels" \
                        f" in {[name for name in data_bundle.datasets.keys() if 'train' not in name]} " \
                        f"data set but not in train data set!."
-            logger.warn(warn_msg)
+            logger.warning(warn_msg)
             print(warn_msg)
         
         has_target_datasets = [dataset for name, dataset in data_bundle.datasets.items() if
diff --git a/fastNLP/io/pipe/utils.py b/fastNLP/io/pipe/utils.py
index aa28af08..05dd3cf4 100644
--- a/fastNLP/io/pipe/utils.py
+++ b/fastNLP/io/pipe/utils.py
@@ -138,7 +138,7 @@ def _indexize(data_bundle, input_field_names='words', target_field_names='target
                        f" in {[name for name in data_bundle.datasets.keys() if 'train' not in name]} " \
                        f"data set but not in train data set!.\n" \
                        f"These label(s) are {tgt_vocab._no_create_word}"
-            logger.warn(warn_msg)
+            logger.warning(warn_msg)
             # log.warning(warn_msg)
         tgt_vocab.index_dataset(*[ds for ds in data_bundle.datasets.values() if ds.has_field(target_field_name)], field_name=target_field_name)
         data_bundle.set_vocab(tgt_vocab, target_field_name)
diff --git a/fastNLP/modules/mix_modules/utils.py b/fastNLP/modules/mix_modules/utils.py
index 21d0f05c..04dab056 100644
--- a/fastNLP/modules/mix_modules/utils.py
+++ b/fastNLP/modules/mix_modules/utils.py
@@ -112,7 +112,7 @@ def _jittor2torch(jittor_var: 'jittor.Var', device: Optional[Union[str, int]] =
     # 如果outputs有_grad键，可以实现求导
     no_gradient = not jittor_var.requires_grad if no_gradient is None else no_gradient
     if no_gradient == False:
-        logger.warn("The result tensor will not keep gradients due to differences between jittor and pytorch.")
+        logger.warning("The result tensor will not keep gradients due to differences between jittor and pytorch.")
     jittor_numpy = jittor_var.numpy()
     if not np.issubdtype(jittor_numpy.dtype, np.inexact):
         no_gradient = True
diff --git a/fastNLP/transformers/torch/configuration_utils.py b/fastNLP/transformers/torch/configuration_utils.py
index 948d9873..26a80377 100644
--- a/fastNLP/transformers/torch/configuration_utils.py
+++ b/fastNLP/transformers/torch/configuration_utils.py
@@ -327,7 +327,7 @@ class PretrainedConfig:
 
         # Deal with gradient checkpointing
         if kwargs.get("gradient_checkpointing", False):
-            logger.warn(
+            logger.warning(
                 "Passing `gradient_checkpointing` to a config initialization is deprecated and will be removed in v5 "
                 "Transformers. Using `model.gradient_checkpointing_enable()` instead, or if you are using the "
                 "`Trainer` API, pass `gradient_checkpointing=True` in your `TrainingArguments`."
diff --git a/fastNLP/transformers/torch/generation_beam_search.py b/fastNLP/transformers/torch/generation_beam_search.py
index 117d9a38..1c3fc592 100644
--- a/fastNLP/transformers/torch/generation_beam_search.py
+++ b/fastNLP/transformers/torch/generation_beam_search.py
@@ -195,7 +195,7 @@ class BeamSearchScorer(BeamScorer):
             )
 
         if "max_length" in kwargs:
-            logger.warn(
+            logger.warning(
                 "Passing `max_length` to BeamSearchScorer is deprecated and has no effect."
                 "`max_length` should be passed directly to `beam_search(...)`, `beam_sample(...)`"
                 ",or `group_beam_search(...)`."
diff --git a/fastNLP/transformers/torch/generation_utils.py b/fastNLP/transformers/torch/generation_utils.py
index 0e6fe5c7..29828c15 100644
--- a/fastNLP/transformers/torch/generation_utils.py
+++ b/fastNLP/transformers/torch/generation_utils.py
@@ -872,7 +872,7 @@ class GenerationMixin:
             max_length = self.config.max_length
         elif max_length is not None and max_new_tokens is not None:
             # Both are set, this is odd, raise a warning
-            logger.warn(
+            logger.warning(
                 "Both `max_length` and `max_new_tokens` have been set but they serve the same purpose.", UserWarning
             )
 
@@ -1239,7 +1239,7 @@ class GenerationMixin:
         logits_processor = logits_processor if logits_processor is not None else LogitsProcessorList()
         stopping_criteria = stopping_criteria if stopping_criteria is not None else StoppingCriteriaList()
         if max_length is not None:
-            logger.warn(
+            logger.warning(
                 "`max_length` is deprecated in this function, use `stopping_criteria=StoppingCriteriaList(MaxLengthCriteria(max_length=max_length))` instead.",
                 UserWarning,
             )
@@ -1475,7 +1475,7 @@ class GenerationMixin:
         logits_processor = logits_processor if logits_processor is not None else LogitsProcessorList()
         stopping_criteria = stopping_criteria if stopping_criteria is not None else StoppingCriteriaList()
         if max_length is not None:
-            logger.warn(
+            logger.warning(
                 "`max_length` is deprecated in this function, use `stopping_criteria=StoppingCriteriaList(MaxLengthCriteria(max_length=max_length))` instead.",
                 UserWarning,
             )
@@ -1726,13 +1726,13 @@ class GenerationMixin:
         logits_processor = logits_processor if logits_processor is not None else LogitsProcessorList()
         stopping_criteria = stopping_criteria if stopping_criteria is not None else StoppingCriteriaList()
         if max_length is not None:
-            logger.warn(
+            logger.warning(
                 "`max_length` is deprecated in this function, use `stopping_criteria=StoppingCriteriaList(MaxLengthCriteria(max_length=max_length))` instead.",
                 UserWarning,
             )
             stopping_criteria = validate_stopping_criteria(stopping_criteria, max_length)
         if len(stopping_criteria) == 0:
-            logger.warn("You don't have defined any stopping_criteria, this will likely loop forever", UserWarning)
+            logger.warning("You don't have defined any stopping_criteria, this will likely loop forever", UserWarning)
         pad_token_id = pad_token_id if pad_token_id is not None else self.config.pad_token_id
         eos_token_id = eos_token_id if eos_token_id is not None else self.config.eos_token_id
         output_scores = output_scores if output_scores is not None else self.config.output_scores
@@ -2030,7 +2030,7 @@ class GenerationMixin:
         logits_processor = logits_processor if logits_processor is not None else LogitsProcessorList()
         stopping_criteria = stopping_criteria if stopping_criteria is not None else StoppingCriteriaList()
         if max_length is not None:
-            logger.warn(
+            logger.warning(
                 "`max_length` is deprecated in this function, use `stopping_criteria=StoppingCriteriaList(MaxLengthCriteria(max_length=max_length))` instead.",
                 UserWarning,
             )
@@ -2325,7 +2325,7 @@ class GenerationMixin:
         logits_processor = logits_processor if logits_processor is not None else LogitsProcessorList()
         stopping_criteria = stopping_criteria if stopping_criteria is not None else StoppingCriteriaList()
         if max_length is not None:
-            logger.warn(
+            logger.warning(
                 "`max_length` is deprecated in this function, use `stopping_criteria=StoppingCriteriaList(MaxLengthCriteria(max_length=max_length))` instead.",
                 UserWarning,
             )
diff --git a/fastNLP/transformers/torch/models/auto/auto_factory.py b/fastNLP/transformers/torch/models/auto/auto_factory.py
index 9eb8ec69..d0969a5b 100644
--- a/fastNLP/transformers/torch/models/auto/auto_factory.py
+++ b/fastNLP/transformers/torch/models/auto/auto_factory.py
@@ -401,7 +401,7 @@ class _BaseAutoModelClass:
                     "the option `trust_remote_code=True` to remove this error."
                 )
             if kwargs.get("revision", None) is None:
-                logger.warn(
+                logger.warning(
                     "Explicitly passing a `revision` is encouraged when loading a model with custom code to ensure "
                     "no malicious code has been contributed in a newer revision."
                 )
diff --git a/fastNLP/transformers/torch/models/auto/configuration_auto.py b/fastNLP/transformers/torch/models/auto/configuration_auto.py
index 45d3c071..1289071d 100644
--- a/fastNLP/transformers/torch/models/auto/configuration_auto.py
+++ b/fastNLP/transformers/torch/models/auto/configuration_auto.py
@@ -130,7 +130,7 @@ class _LazyLoadAllMappings(OrderedDict):
     def _initialize(self):
         if self._initialized:
             return
-        # logger.warn(
+        # logger.warning(
         #     "ALL_PRETRAINED_CONFIG_ARCHIVE_MAP is deprecated and will be removed in v5 of Transformers. "
         #     "It does not contain all available model checkpoints, far from it. Checkout hf.co/models for that.",
         #     FutureWarning,
diff --git a/fastNLP/transformers/torch/models/auto/modeling_auto.py b/fastNLP/transformers/torch/models/auto/modeling_auto.py
index aace27a2..dbf4b610 100644
--- a/fastNLP/transformers/torch/models/auto/modeling_auto.py
+++ b/fastNLP/transformers/torch/models/auto/modeling_auto.py
@@ -306,7 +306,7 @@ AutoModelForSpeechSeq2Seq = auto_class_update(
 class AutoModelWithLMHead(_AutoModelWithLMHead):
     @classmethod
     def from_config(cls, config):
-        logger.warn(
+        logger.warning(
             "The class `AutoModelWithLMHead` is deprecated and will be removed in a future version. Please use "
             "`AutoModelForCausalLM` for causal language models, `AutoModelForMaskedLM` for masked language models and "
             "`AutoModelForSeq2SeqLM` for encoder-decoder models.",
@@ -316,7 +316,7 @@ class AutoModelWithLMHead(_AutoModelWithLMHead):
 
     @classmethod
     def from_pretrained(cls, pretrained_model_name_or_path, *model_args, **kwargs):
-        logger.warn(
+        logger.warning(
             "The class `AutoModelWithLMHead` is deprecated and will be removed in a future version. Please use "
             "`AutoModelForCausalLM` for causal language models, `AutoModelForMaskedLM` for masked language models and "
             "`AutoModelForSeq2SeqLM` for encoder-decoder models.",
diff --git a/fastNLP/transformers/torch/models/bart/modeling_bart.py b/fastNLP/transformers/torch/models/bart/modeling_bart.py
index 7219f49a..377afa41 100644
--- a/fastNLP/transformers/torch/models/bart/modeling_bart.py
+++ b/fastNLP/transformers/torch/models/bart/modeling_bart.py
@@ -513,7 +513,7 @@ class BartPretrainedModel(PreTrainedModel):
 
 class PretrainedBartModel(BartPretrainedModel):
     def __init_subclass__(self):
-        logger.warn(
+        logger.warning(
             "The class `PretrainedBartModel` has been depreciated, please use `BartPretrainedModel` instead.",
             FutureWarning,
         )
diff --git a/fastNLP/transformers/torch/models/bert/modeling_bert.py b/fastNLP/transformers/torch/models/bert/modeling_bert.py
index b95da0df..79f1c459 100644
--- a/fastNLP/transformers/torch/models/bert/modeling_bert.py
+++ b/fastNLP/transformers/torch/models/bert/modeling_bert.py
@@ -1374,7 +1374,7 @@ class BertForNextSentencePrediction(BertPreTrainedModel):
         """
 
         if "next_sentence_label" in kwargs:
-            logger.warn(
+            logger.warning(
                 "The `next_sentence_label` argument is deprecated and will be removed in a future version, use `labels` instead.",
                 FutureWarning,
             )
diff --git a/fastNLP/transformers/torch/models/cpt/modeling_cpt.py b/fastNLP/transformers/torch/models/cpt/modeling_cpt.py
index 2910cc26..df7d477b 100644
--- a/fastNLP/transformers/torch/models/cpt/modeling_cpt.py
+++ b/fastNLP/transformers/torch/models/cpt/modeling_cpt.py
@@ -724,7 +724,7 @@ class CPTDecoder(CPTPretrainedModel):
             if getattr(self.config, "gradient_checkpointing", False) and self.training:
 
                 if use_cache:
-                    logger.warn(
+                    logger.warning(
                         "`use_cache=True` is incompatible with `config.gradient_checkpointing=True`. Setting "
                         "`use_cache=False`..."
                     )
diff --git a/fastNLP/transformers/torch/tokenization_utils_base.py b/fastNLP/transformers/torch/tokenization_utils_base.py
index 3a033c96..a04dbaf1 100644
--- a/fastNLP/transformers/torch/tokenization_utils_base.py
+++ b/fastNLP/transformers/torch/tokenization_utils_base.py
@@ -312,7 +312,7 @@ class BatchEncoding(UserDict):
         """
         if not self._encodings:
             raise ValueError("words() is not available when using Python-based tokenizers")
-        logger.warn(
+        logger.warning(
             "`BatchEncoding.words()` property is deprecated and should be replaced with the identical, "
             "but more self-explanatory `BatchEncoding.word_ids()` property.",
             FutureWarning,
@@ -1601,7 +1601,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
                     f"Calling {cls.__name__}.from_pretrained() with the path to a single file or url is not "
                     "supported for this tokenizer. Use a model identifier or the path to a directory instead."
                 )
-            logger.warn(
+            logger.warning(
                 f"Calling {cls.__name__}.from_pretrained() with the path to a single file or url is deprecated and "
                 "won't be possible anymore in v5. Use a model identifier or the path to a directory instead.",
                 FutureWarning,
@@ -2163,7 +2163,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
         # Get padding strategy
         if padding is False and old_pad_to_max_length:
             if verbose:
-                logger.warn(
+                logger.warning(
                     "The `pad_to_max_length` argument is deprecated and will be removed in a future version, "
                     "use `padding=True` or `padding='longest'` to pad to the longest sequence in the batch, or "
                     "use `padding='max_length'` to pad to a max length. In this case, you can give a specific "
@@ -2184,7 +2184,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
                             "To pad to max length, use `padding='max_length'`."
                         )
                     if old_pad_to_max_length is not False:
-                        logger.warn("Though `pad_to_max_length` = `True`, it is ignored because `padding`=`True`.")
+                        logger.warning("Though `pad_to_max_length` = `True`, it is ignored because `padding`=`True`.")
                 padding_strategy = PaddingStrategy.LONGEST  # Default to pad to the longest sequence in the batch
             elif not isinstance(padding, PaddingStrategy):
                 padding_strategy = PaddingStrategy(padding)
@@ -2196,7 +2196,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
         # Get truncation strategy
         if truncation is False and old_truncation_strategy != "do_not_truncate":
             if verbose:
-                logger.warn(
+                logger.warning(
                     "The `truncation_strategy` argument is deprecated and will be removed in a future version, "
                     "use `truncation=True` to truncate examples to a max length. You can give a specific "
                     "length with `max_length` (e.g. `max_length=45`) or leave max_length to None to truncate to the "
@@ -3352,7 +3352,7 @@ model_inputs["labels"] = labels["input_ids"]
 See the documentation of your specific tokenizer for more details on the specific arguments to the tokenizer of choice.
 For a more complete example, see the implementation of `prepare_seq2seq_batch`.
 """
-        logger.warn(formatted_warning, FutureWarning)
+        logger.warning(formatted_warning, FutureWarning)
         # mBART-specific kwargs that should be ignored by other models.
         kwargs.pop("src_lang", None)
         kwargs.pop("tgt_lang", None)