Removing Sequence and Token classification models. Removing integrati…

…on tests for now

src/transformers/models/auto/modeling_auto.py

@@ -1287,7 +1287,6 @@ class _BaseModelWithGenerate(PreTrainedModel, GenerationMixin):
         ("tapas", "TapasForSequenceClassification"),
         ("transfo-xl", "TransfoXLForSequenceClassification"),
         ("umt5", "UMT5ForSequenceClassification"),
-        ("vaultgemma", "VaultGemmaForSequenceClassification"),
         ("xlm", "XLMForSequenceClassification"),
         ("xlm-roberta", "XLMRobertaForSequenceClassification"),
         ("xlm-roberta-xl", "XLMRobertaXLForSequenceClassification"),
@@ -1493,7 +1492,6 @@ class _BaseModelWithGenerate(PreTrainedModel, GenerationMixin):
         ("t5", "T5ForTokenClassification"),
         ("t5gemma", "T5GemmaForTokenClassification"),
         ("umt5", "UMT5ForTokenClassification"),
-        ("vaultgemma", "VaultGemmaForTokenClassification"),
         ("xlm", "XLMForTokenClassification"),
         ("xlm-roberta", "XLMRobertaForTokenClassification"),
         ("xlm-roberta-xl", "XLMRobertaXLForTokenClassification"),

src/transformers/models/vaultgemma/modeling_vaultgemma.py

@@ -29,11 +29,7 @@
 from ...generation import GenerationMixin
 from ...masking_utils import create_causal_mask, create_sliding_window_causal_mask
 from ...modeling_flash_attention_utils import FlashAttentionKwargs
-from ...modeling_layers import (
-    GenericForSequenceClassification,
-    GenericForTokenClassification,
-    GradientCheckpointingLayer,
-)
+from ...modeling_layers import GradientCheckpointingLayer
 from ...modeling_outputs import BaseModelOutputWithPast, CausalLMOutputWithPast
 from ...modeling_rope_utils import ROPE_INIT_FUNCTIONS, dynamic_rope_update
 from ...modeling_utils import ALL_ATTENTION_FUNCTIONS, PreTrainedModel
@@ -572,18 +568,4 @@ def forward(
         )
 
 
-class VaultGemmaForSequenceClassification(GenericForSequenceClassification, VaultGemmaPreTrainedModel):
-    pass
-
-
-class VaultGemmaForTokenClassification(GenericForTokenClassification, VaultGemmaPreTrainedModel):
-    pass
-
-
-__all__ = [
-    "VaultGemmaForCausalLM",
-    "VaultGemmaModel",
-    "VaultGemmaPreTrainedModel",
-    "VaultGemmaForSequenceClassification",
-    "VaultGemmaForTokenClassification",
-]
+__all__ = ["VaultGemmaForCausalLM", "VaultGemmaModel", "VaultGemmaPreTrainedModel"]

src/transformers/models/vaultgemma/modular_vaultgemma.py

@@ -23,8 +23,6 @@
     Gemma2Attention,
     Gemma2DecoderLayer,
     Gemma2ForCausalLM,
-    Gemma2ForSequenceClassification,
-    Gemma2ForTokenClassification,
     Gemma2MLP,
     Gemma2Model,
     Gemma2PreTrainedModel,
@@ -112,19 +110,10 @@ class VaultGemmaForCausalLM(Gemma2ForCausalLM):
     pass
 
 
-class VaultGemmaForSequenceClassification(Gemma2ForSequenceClassification):
-    pass
-
-
-class VaultGemmaForTokenClassification(Gemma2ForTokenClassification):
-    pass
-
 
 __all__ = [
     "VaultGemmaConfig",
     "VaultGemmaForCausalLM",
     "VaultGemmaModel",
     "VaultGemmaPreTrainedModel",
-    "VaultGemmaForSequenceClassification",
-    "VaultGemmaForTokenClassification",
 ]