Update modeling_hyperclovax.py

modeling_hyperclovax.py

@@ -43,7 +43,7 @@ from transformers.modeling_rope_utils import ROPE_INIT_FUNCTIONS, dynamic_rope_u
 from transformers.modeling_utils import ALL_ATTENTION_FUNCTIONS, PreTrainedModel
 from transformers.processing_utils import Unpack
 from transformers.pytorch_utils import ALL_LAYERNORM_LAYERS
-from transformers.utils import LossKwargs, auto_docstring, can_return_tuple, is_torch_flex_attn_available, logging
+from transformers.utils import TransformersKwargs, auto_docstring, can_return_tuple, is_torch_flex_attn_available, logging
 from .configuration_hyperclovax import HyperCLOVAXConfig
 if is_torch_flex_attn_available():
     from torch.nn.attention.flex_attention import BlockMask
@@ -620,7 +620,7 @@ class HyperCLOVAXModel(HyperCLOVAXPreTrainedModel):
         return causal_mask
 
 
-class KwargsForCausalLM(FlashAttentionKwargs, LossKwargs): ...
+class KwargsForCausalLM(FlashAttentionKwargs, TransformersKwargs): ...
 
 
 @auto_docstring