works

config.json

@@ -1,15 +1,31 @@
 {
+  "_name_or_path": "nezha-cn-base",
+  "architectures": [
+    "NeZhaForMaskedLM"
+  ],
   "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 2,
+  "classifier_dropout": 0.1,
+  "embedding_size": 128,
+  "eos_token_id": 3,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
   "initializer_range": 0.02,
+  "inner_group_num": 1,
   "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
   "max_position_embeddings": 512,
   "max_relative_position": 64,
+  "model_type": "nezha",
   "num_attention_heads": 12,
+  "num_hidden_groups": 1,
   "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "torch_dtype": "float32",
+  "transformers_version": "4.20.0.dev0",
   "type_vocab_size": 2,
-  "vocab_size": 21128,
-  "use_relative_position": true
+  "use_cache": true,
+  "use_relative_position": true,
+  "vocab_size": 21128
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d188d91259e4e5fd6aafdf54ef95607183507f9e1e6a5c536b790a7ee0bc4a4e
-size 410021031
+oid sha256:8a11b83560edb01f0eec1470a5abc84a743a51a766558683233fa06d5896cd8e
+size 410022183

tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "name_or_path": "nezha-cn-base",
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": null,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}