change config name

config.json

@@ -1,6 +1,6 @@
 {
   "auto_map": {
-    "AutoConfig": "configuration_bert.BertConfig",
+    "AutoConfig": "configuration_bert.XLMFlashConfig",
     "AutoModel": "modeling_bert.BertModel",
     "AutoModelForPreTraining": "modeling_bert.BertForPreTraining",
     "AutoModelForMaskedLM": "modeling_bert.BertForPreTraining"

configuration_bert.py

@@ -1,6 +1,6 @@
 from transformers import PretrainedConfig
 
-class BertConfig(PretrainedConfig):
+class XLMFlashConfig(PretrainedConfig):
     def __init__(
             self,
             vocab_size=30522,

modeling_bert.py

@@ -19,7 +19,7 @@ import torch
 import torch.nn as nn
 import torch.nn.functional as F
 from einops import rearrange
-from transformers import BertConfig, PretrainedConfig, XLMRobertaConfig # TODO check whether to use XLMRobertaConfig
+from transformers import BertConfig, PretrainedConfig
 from transformers.modeling_utils import PreTrainedModel
 from transformers.models.bert.modeling_bert import (
     BaseModelOutputWithPoolingAndCrossAttentions,
@@ -32,6 +32,7 @@ from .bert_padding import (
     pad_input,
     unpad_input,
 )
+from .configuration_bert import XLMFlashConfig
 from .block import Block
 from .embedding import BertEmbeddings
 from .mha import MHA
@@ -345,7 +346,7 @@ class BertPreTrainedModel(PreTrainedModel):
     """An abstract class to handle weights initialization and
     a simple interface for dowloading and loading pretrained models.
     """
-    config_class = XLMRobertaConfig
+    config_class = XLMFlashConfig
     base_model_prefix = "bert"
     supports_gradient_checkpointing = True