Replace whole-word `LossKwargs` with `TransformersKwargs` in modeling*.py

modeling_opensci.py

@@ -40,7 +40,7 @@ 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,
+    TransformersKwargs,
     add_code_sample_docstrings,
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
@@ -750,7 +750,7 @@ class OpensciModel(OpensciPreTrainedModel):
         return causal_mask
 
 
-class KwargsForCausalLM(FlashAttentionKwargs, LossKwargs): ...
+class KwargsForCausalLM(FlashAttentionKwargs, TransformersKwargs): ...
 
 
 class OpensciForCausalLM(OpensciPreTrainedModel, GenerationMixin):