fix bug in .py

configuration_minicpm.py

@@ -174,7 +174,6 @@ class MiniCPM3Config(PretrainedConfig):
         self.use_cache = use_cache
         self.rope_theta = rope_theta
         self.rope_scaling = rope_scaling
-        self._rope_scaling_validation()
         self.attention_bias = attention_bias
         self.attention_dropout = attention_dropout
         self.scale_emb = scale_emb

modeling_minicpm.py

@@ -48,7 +48,7 @@ from transformers.utils import (
     replace_return_docstrings,
 )
 from transformers.utils.import_utils import is_torch_fx_available
-from .configuration_minicpm import MiniCPMConfig
+from .configuration_minicpm import MiniCPM3Config
 import re
 
 try:
@@ -69,7 +69,7 @@ if is_torch_fx_available():
 
 logger = logging.get_logger(__name__)
 
-_CONFIG_FOR_DOC = "MiniCPMConfig"
+_CONFIG_FOR_DOC = "MiniCPM3Config"
 
 
 def _get_unpad_data(attention_mask):
@@ -331,7 +331,7 @@ def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor:
 class MiniCPMAttention(nn.Module):
     """Multi-headed attention from 'Attention Is All You Need' paper"""
 
-    def __init__(self, config: MiniCPMConfig, layer_idx: Optional[int] = None):
+    def __init__(self, config: MiniCPM3Config, layer_idx: Optional[int] = None):
         super().__init__()
         self.config = config
         self.layer_idx = layer_idx
@@ -784,7 +784,7 @@ class MiniCPMSdpaAttention(MiniCPMAttention):
         if output_attentions:
             # TODO: Improve this warning with e.g. `model.config.attn_implementation = "manual"` once this is implemented.
             logger.warning_once(
-                "MiniCPMModel is using MiniCPMSdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True`. Falling back to the manual attention implementation, "
+                "MiniCPM3Model is using MiniCPMSdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True`. Falling back to the manual attention implementation, "
                 'but specifying the manual implementation will be required from Transformers version v5.0.0 onwards. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.'
             )
             return super().forward(
@@ -884,7 +884,7 @@ MINICPM_ATTENTION_CLASSES = {
 
 
 class MiniCPMDecoderLayer(nn.Module):
-    def __init__(self, config: MiniCPMConfig, layer_idx: int):
+    def __init__(self, config: MiniCPM3Config, layer_idx: int):
         super().__init__()
         self.hidden_size = config.hidden_size
         self.self_attn = MINICPM_ATTENTION_CLASSES[config._attn_implementation](config=config, layer_idx=layer_idx)
@@ -968,7 +968,7 @@ MINICPM_START_DOCSTRING = r"""
     and behavior.
 
     Parameters:
-        config ([`MiniCPMConfig`]):
+        config ([`MiniCPM3Config`]):
             Model configuration class with all the parameters of the model. Initializing with a config file does not
             load the weights associated with the model, only the configuration. Check out the
             [`~PreTrainedModel.from_pretrained`] method to load the model weights.
@@ -980,7 +980,7 @@ MINICPM_START_DOCSTRING = r"""
     MINICPM_START_DOCSTRING,
 )
 class MiniCPM3PreTrainedModel(PreTrainedModel):
-    config_class = MiniCPMConfig
+    config_class = MiniCPM3Config
     base_model_prefix = "model"
     supports_gradient_checkpointing = True
     _no_split_modules = ["MiniCPMDecoderLayer"]
@@ -1080,10 +1080,10 @@ class MiniCPM3Model(MiniCPM3PreTrainedModel):
     Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`MiniCPMDecoderLayer`]
 
     Args:
-        config: MiniCPMConfig
+        config: MiniCPM3Config
     """
 
-    def __init__(self, config: MiniCPMConfig):
+    def __init__(self, config: MiniCPM3Config):
         super().__init__(config)
         self.padding_idx = config.pad_token_id
         self.vocab_size = config.vocab_size
@@ -1244,7 +1244,7 @@ class MiniCPM3ForCausalLM(MiniCPM3PreTrainedModel):
 
     def __init__(self, config):
         super().__init__(config)
-        self.model = MiniCPMModel(config)
+        self.model = MiniCPM3Model(config)
         self.vocab_size = config.vocab_size
         self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
 
@@ -1469,7 +1469,7 @@ class MiniCPM3ForSequenceClassification(MiniCPM3PreTrainedModel):
     def __init__(self, config):
         super().__init__(config)
         self.num_labels = config.num_labels
-        self.model = MiniCPMModel(config)
+        self.model = MiniCPM3Model(config)
         self.score = nn.Linear(config.hidden_size, self.num_labels, bias=False)
 
         # Initialize weights and apply final processing