Remove more unused attributes in config classes (#20858)

Remove more unused attributes in config classes
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

src/transformers/models/ctrl/configuration_ctrl.py

@@ -52,8 +52,6 @@ class CTRLConfig(PretrainedConfig):
             The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
         embd_pdrop (`int`, *optional*, defaults to 0.1):
             The dropout ratio for the embeddings.
-        attn_pdrop (`float`, *optional*, defaults to 0.1):
-            The dropout ratio for the attention.
         layer_norm_epsilon (`float`, *optional*, defaults to 1e-6):
             The epsilon to use in the layer normalization layers
         initializer_range (`float`, *optional*, defaults to 0.02):
@@ -96,7 +94,6 @@ class CTRLConfig(PretrainedConfig):
         n_head=16,
         resid_pdrop=0.1,
         embd_pdrop=0.1,
-        attn_pdrop=0.1,
         layer_norm_epsilon=1e-6,
         initializer_range=0.02,
         summary_type="cls_index",
@@ -115,7 +112,6 @@ class CTRLConfig(PretrainedConfig):
         self.dff = dff
         self.resid_pdrop = resid_pdrop
         self.embd_pdrop = embd_pdrop
-        self.attn_pdrop = attn_pdrop
         self.layer_norm_epsilon = layer_norm_epsilon
         self.initializer_range = initializer_range
 

src/transformers/models/data2vec/configuration_data2vec_vision.py

@@ -40,9 +40,6 @@ class Data2VecVisionConfig(PretrainedConfig):
     [facebook/data2vec-vision-base](https://huggingface.co/facebook/data2vec-vision-base) architecture.
 
     Args:
-        vocab_size (`int`, *optional*, defaults to 8092):
-            Vocabulary size of the Data2VecVision model. Defines the number of different image tokens that can be used
-            during pre-training.
         hidden_size (`int`, *optional*, defaults to 768):
             Dimensionality of the encoder layers and the pooler layer.
         num_hidden_layers (`int`, *optional*, defaults to 12):
@@ -118,7 +115,6 @@ class Data2VecVisionConfig(PretrainedConfig):
 
     def __init__(
         self,
-        vocab_size=8192,
         hidden_size=768,
         num_hidden_layers=12,
         num_attention_heads=12,
@@ -150,7 +146,6 @@ class Data2VecVisionConfig(PretrainedConfig):
     ):
         super().__init__(**kwargs)
 
-        self.vocab_size = vocab_size
         self.hidden_size = hidden_size
         self.num_hidden_layers = num_hidden_layers
         self.num_attention_heads = num_attention_heads

src/transformers/models/esm/configuration_esm.py

@@ -59,9 +59,6 @@ class EsmConfig(PretrainedConfig):
             Number of attention heads for each attention layer in the Transformer encoder.
         intermediate_size (`int`, *optional*, defaults to 3072):
             Dimensionality of the "intermediate" (often named feed-forward) layer in the Transformer encoder.
-        hidden_act (`str` or `Callable`, *optional*, defaults to `"gelu"`):
-            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
-            `"relu"`, `"silu"` and `"gelu_new"` are supported.
         hidden_dropout_prob (`float`, *optional*, defaults to 0.1):
             The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
         attention_probs_dropout_prob (`float`, *optional*, defaults to 0.1):
@@ -113,7 +110,6 @@ class EsmConfig(PretrainedConfig):
         num_hidden_layers=12,
         num_attention_heads=12,
         intermediate_size=3072,
-        hidden_act="gelu",
         hidden_dropout_prob=0.1,
         attention_probs_dropout_prob=0.1,
         max_position_embeddings=1026,
@@ -135,7 +131,6 @@ class EsmConfig(PretrainedConfig):
         self.hidden_size = hidden_size
         self.num_hidden_layers = num_hidden_layers
         self.num_attention_heads = num_attention_heads
-        self.hidden_act = hidden_act
         self.intermediate_size = intermediate_size
         self.hidden_dropout_prob = hidden_dropout_prob
         self.attention_probs_dropout_prob = attention_probs_dropout_prob

src/transformers/models/groupvit/configuration_groupvit.py

@@ -221,7 +221,6 @@ class GroupViTVisionConfig(PretrainedConfig):
         initializer_factor=1.0,
         assign_eps=1.0,
         assign_mlp_ratio=[0.5, 4],
-        qkv_bias=True,
         **kwargs
     ):
         super().__init__(**kwargs)
@@ -249,7 +248,6 @@ class GroupViTVisionConfig(PretrainedConfig):
         self.initializer_factor = initializer_factor
         self.assign_eps = assign_eps
         self.assign_mlp_ratio = assign_mlp_ratio
-        self.qkv_bias = qkv_bias
 
     @classmethod
     def from_pretrained(cls, pretrained_model_name_or_path: Union[str, os.PathLike], **kwargs) -> "PretrainedConfig":