Upload 14 files

config.json

@@ -0,0 +1,36 @@
+{
+  "_name_or_path": "iioSnail/ChineseBERT-for-csc",
+  "architectures": [
+    "ChineseBertForCSC"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "auto_map": {
+    "AutoModel": "csc_model.ChineseBertForCSC"
+  },
+  "classifier_dropout": null,
+  "directionality": "bidi",
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.27.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 23236
+}

config/STFANGSO.TTF24.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09e11244c473c9272b6e9d2b81aaead7d3adbeb9c4c6fde5c5fe495401ceb065
+size 107071616

config/STXINGKA.TTF24.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:531e568202ceff79cc621775050b36c1c627463fc535fa3231a76c904b886bfd
+size 107071616

config/pinyin2tensor.json

@@ -0,0 +1 @@
+{"ling2": [17, 14, 19, 12, 2, 0, 0, 0], "yuan2": [30, 26, 6, 19, 2, 0, 0, 0], "xing1": [29, 14, 19, 12, 1, 0, 0, 0], "yi1": [30, 14, 1, 0, 0, 0, 0, 0], "yi2": [30, 14, 2, 0, 0, 0, 0, 0], "yi4": [30, 14, 4, 0, 0, 0, 0, 0], "ding1": [9, 14, 19, 12, 1, 0, 0, 0], "zheng1": [31, 13, 10, 19, 12, 1, 0, 0], "qi1": [22, 14, 1, 0, 0, 0, 0, 0], "qi2": [22, 14, 2, 0, 0, 0, 0, 0], "wan4": [28, 6, 19, 4, 0, 0, 0, 0], "mo4": [18, 20, 4, 0, 0, 0, 0, 0], "zhang4": [31, 13, 6, 19, 12, 4, 0, 0], "san1": [24, 6, 19, 1, 0, 0, 0, 0], "shang4": [24, 13, 6, 19, 12, 4, 0, 0], "shang3": [24, 13, 6, 19, 12, 3, 0, 0], "xia4": [29, 14, 6, 4, 0, 0, 0, 0], "bu4": [7, 26, 4, 0, 0, 0, 0, 0], "fou3": [11, 20, 26, 3, 0, 0, 0, 0], "fou1": [11, 20, 26, 1, 0, 0, 0, 0], "fu1": [11, 26, 1, 0, 0, 0, 0, 0], "bu2": [7, 26, 2, 0, 0, 0, 0, 0], "yu3": [30, 26, 3, 0, 0, 0, 0, 0], "yu2": [30, 26, 2, 0, 0, 0, 0, 0], "yu4": [30, 26, 4, 0, 0, 0, 0, 0], "gai4": [12, 6, 14, 4, 0, 0, 0, 0], "chou3": [8, 13, 20, 26, 3, 0, 0, 0], "zhuan1": [31, 13, 26, 6, 19, 1, 0, 0], "qie3": [22, 14, 10, 3, 0, 0, 0, 0], "ju1": [15, 26, 1, 0, 0, 0, 0, 0], "cu2": [8, 26, 2, 0, 0, 0, 0, 0], "pi1": [21, 14, 1, 0, 0, 0, 0, 0], "shi4": [24, 13, 14, 4, 0, 0, 0, 0], "qiu1": [22, 14, 26, 1, 0, 0, 0, 0], "bing3": [7, 14, 19, 12, 3, 0, 0, 0], "bing4": [7, 14, 19, 12, 4, 0, 0, 0], "ye4": [30, 10, 4, 0, 0, 0, 0, 0], "cong2": [8, 20, 19, 12, 2, 0, 0, 0], "dong1": [9, 20, 19, 12, 1, 0, 0, 0], "si1": [24, 14, 1, 0, 0, 0, 0, 0], "cheng2": [8, 13, 10, 19, 12, 2, 0, 0], "sheng4": [24, 13, 10, 19, 12, 4, 0, 0], "zheng3": [31, 13, 10, 19, 12, 3, 0, 0], "diu1": [9, 14, 26, 1, 0, 0, 0, 0], "liang3": [17, 14, 6, 19, 12, 3, 0, 0], "yan2": [30, 6, 19, 2, 0, 0, 0, 0], "ban4": [7, 6, 19, 4, 0, 0, 0, 0], "bang4": [7, 6, 19, 12, 4, 0, 0, 0], "sang4": [24, 6, 19, 12, 4, 0, 0, 0], "sang1": [24, 6, 19, 12, 1, 0, 0, 0], "gun3": [12, 26, 19, 3, 0, 0, 0, 0], "ge4": [12, 10, 4, 0, 0, 0, 0, 0], "gan4": [12, 6, 19, 4, 0, 0, 0, 0], "ge3": [12, 10, 3, 0, 0, 0, 0, 0], "ya1": [30, 6, 1, 0, 0, 0, 0, 0], "zhong1": [31, 13, 20, 19, 12, 1, 0, 0], "zhong4": [31, 13, 20, 19, 12, 4, 0, 0], "feng1": [11, 10, 19, 12, 1, 0, 0, 0], "chuan4": [8, 13, 26, 6, 19, 4, 0, 0], "guan4": [12, 26, 6, 19, 4, 0, 0, 0], "quan4": [22, 26, 6, 19, 4, 0, 0, 0], "lin2": [17, 14, 19, 2, 0, 0, 0, 0], "zhu3": [31, 13, 26, 3, 0, 0, 0, 0], "wan2": [28, 6, 19, 2, 0, 0, 0, 0], "dan1": [9, 6, 19, 1, 0, 0, 0, 0], "wei4": [28, 10, 14, 4, 0, 0, 0, 0], "wei2": [28, 10, 14, 2, 0, 0, 0, 0], "zhu4": [31, 13, 26, 4, 0, 0, 0, 0], "jing3": [15, 14, 19, 12, 3, 0, 0, 0], "dan3": [9, 6, 19, 3, 0, 0, 0, 0], "li4": [17, 14, 4, 0, 0, 0, 0, 0], "li2": [17, 14, 2, 0, 0, 0, 0, 0], "ju3": [15, 26, 3, 0, 0, 0, 0, 0], "pie3": [21, 14, 10, 3, 0, 0, 0, 0], "ai4": [6, 14, 4, 0, 0, 0, 0, 0], "nai3": [19, 6, 14, 3, 0, 0, 0, 0], "ai3": [6, 14, 3, 0, 0, 0, 0, 0], "jiu3": [15, 14, 26, 3, 0, 0, 0, 0], "me": [18, 10, 5, 0, 0, 0, 0, 0], "yao1": [30, 6, 20, 1, 0, 0, 0, 0], "mo2": [18, 20, 2, 0, 0, 0, 0, 0], "ma": [18, 6, 5, 0, 0, 0, 0, 0], "zhi1": [31, 13, 14, 1, 0, 0, 0, 0], "zhu1": [31, 13, 26, 1, 0, 0, 0, 0], "zhi4": [31, 13, 14, 4, 0, 0, 0, 0], "wu1": [28, 26, 1, 0, 0, 0, 0, 0], "wu4": [28, 26, 4, 0, 0, 0, 0, 0], "zha4": [31, 13, 6, 4, 0, 0, 0, 0], "zuo4": [31, 26, 20, 4, 0, 0, 0, 0], "hu1": [13, 26, 1, 0, 0, 0, 0, 0], "fa2": [11, 6, 2, 0, 0, 0, 0, 0], "le4": [17, 10, 4, 0, 0, 0, 0, 0], "yue4": [30, 26, 10, 4, 0, 0, 0, 0], "ping1": [21, 14, 19, 12, 1, 0, 0, 0], "pang1": [21, 6, 19, 12, 1, 0, 0, 0], "qiao2": [22, 14, 6, 20, 2, 0, 0, 0], "guai1": [12, 26, 6, 14, 1, 0, 0, 0], "yi3": [30, 14, 3, 0, 0, 0, 0, 0], "jue2": [15, 26, 10, 2, 0, 0, 0, 0], "mie1": [18, 14, 10, 1, 0, 0, 0, 0], "nie4": [19, 14, 10, 4, 0, 0, 0, 0], "jiu1": [15, 14, 26, 1, 0, 0, 0, 0], "qi3": [22, 14, 3, 0, 0, 0, 0, 0], "qi4": [22, 14, 4, 0, 0, 0, 0, 0], "ye3": [30, 10, 3, 0, 0, 0, 0, 0], "xi2": [29, 14, 2, 0, 0, 0, 0, 0], "xiang1": [29, 14, 6, 19, 12, 1, 0, 0], "shu1": [24, 13, 26, 1, 0, 0, 0, 0], "ji1": [15, 14, 1, 0, 0, 0, 0, 0], "mai3": [18, 6, 14, 3, 0, 0, 0, 0], "luan4": [17, 26, 6, 19, 4, 0, 0, 0], "ru3": [23, 26, 3, 0, 0, 0, 0, 0], "qian2": [22, 14, 6, 19, 2, 0, 0, 0], "gan1": [12, 6, 19, 1, 0, 0, 0, 0], "gui1": [12, 26, 14, 1, 0, 0, 0, 0], "le": [17, 10, 5, 0, 0, 0, 0, 0], "liao3": [17, 14, 6, 20, 3, 0, 0, 0], "liao4": [17, 14, 6, 20, 4, 0, 0, 0], "zi4": [31, 14, 4, 0, 0, 0, 0, 0], "er4": [10, 23, 4, 0, 0, 0, 0, 0], "yu1": [30, 26, 1, 0, 0, 0, 0, 0], "xu1": [29, 26, 1, 0, 0, 0, 0, 0], "kui1": [16, 26, 14, 1, 0, 0, 0, 0], "yun2": [30, 26, 19, 2, 0, 0, 0, 0], "hu4": [13, 26, 4, 0, 0, 0, 0, 0], "wu3": [28, 26, 3, 0, 0, 0, 0, 0], "jing4": [15, 14, 19, 12, 4, 0, 0, 0], "gen4": [12, 10, 19, 4, 0, 0, 0, 0], "xuan1": [29, 26, 6, 19, 1, 0, 0, 0], "geng4": [12, 10, 19, 12, 4, 0, 0, 0], "ya4": [30, 6, 4, 0, 0, 0, 0, 0], "xie1": [29, 14, 10, 1, 0, 0, 0, 0], "suo4": [24, 26, 20, 4, 0, 0, 0, 0], "suo1": [24, 26, 20, 1, 0, 0, 0, 0], "e4": [10, 4, 0, 0, 0, 0, 0, 0], "ji2": [15, 14, 2, 0, 0, 0, 0, 0], "wang2": [28, 6, 19, 12, 2, 0, 0, 0], "wu2": [28, 26, 2, 0, 0, 0, 0, 0], "kang4": [16, 6, 19, 12, 4, 0, 0, 0], "gang1": [12, 6, 19, 12, 1, 0, 0, 0], "geng1": [12, 10, 19, 12, 1, 0, 0, 0], "jiao1": [15, 14, 6, 20, 1, 0, 0, 0], "hai4": [13, 6, 14, 4, 0, 0, 0, 0], "jie1": [15, 14, 10, 1, 0, 0, 0, 0], "chan3": [8, 13, 6, 19, 3, 0, 0, 0], "heng1": [13, 10, 19, 12, 1, 0, 0, 0], "xiang3": [29, 14, 6, 19, 12, 3, 0, 0], "peng1": [21, 10, 19, 12, 1, 0, 0, 0], "mu3": [18, 26, 3, 0, 0, 0, 0, 0], "jing1": [15, 14, 19, 12, 1, 0, 0, 0], "ting2": [25, 14, 19, 12, 2, 0, 0, 0], "liang4": [17, 14, 6, 19, 12, 4, 0, 0], "liang2": [17, 14, 6, 19, 12, 2, 0, 0], "qin1": [22, 14, 19, 1, 0, 0, 0, 0], "qing4": [22, 14, 19, 12, 4, 0, 0, 0], "bo2": [7, 20, 2, 0, 0, 0, 0, 0], "xie4": [29, 14, 10, 4, 0, 0, 0, 0], "ren2": [23, 10, 19, 2, 0, 0, 0, 0], "shen2": [24, 13, 10, 19, 2, 0, 0, 0], "shi2": [24, 13, 14, 2, 0, 0, 0, 0], "ding3": [9, 14, 19, 12, 3, 0, 0, 0], "ze4": [31, 10, 4, 0, 0, 0, 0, 0], "jin3": [15, 14, 19, 3, 0, 0, 0, 0], "fu4": [11, 26, 4, 0, 0, 0, 0, 0], "nu2": [19, 26, 2, 0, 0, 0, 0, 0], "jin4": [15, 14, 19, 4, 0, 0, 0, 0], "pu1": [21, 26, 1, 0, 0, 0, 0, 0], "pu2": [21, 26, 2, 0, 0, 0, 0, 0], "chou2": [8, 13, 20, 26, 2, 0, 0, 0], "qiu2": [22, 14, 26, 2, 0, 0, 0, 0], "jin1": [15, 14, 19, 1, 0, 0, 0, 0], "jie4": [15, 14, 10, 4, 0, 0, 0, 0], "reng2": [23, 10, 19, 12, 2, 0, 0, 0], "cong1": [8, 20, 19, 12, 1, 0, 0, 0], "fo2": [11, 20, 2, 0, 0, 0, 0, 0], "lun2": [17, 26, 19, 2, 0, 0, 0, 0], "cang1": [8, 6, 19, 12, 1, 0, 0, 0], "zai3": [31, 6, 14, 3, 0, 0, 0, 0], "zi3": [31, 14, 3, 0, 0, 0, 0, 0], "zi1": [31, 14, 1, 0, 0, 0, 0, 0], "ta1": [25, 6, 1, 0, 0, 0, 0, 0], "tuo2": [25, 26, 20, 2, 0, 0, 0, 0], "xian1": [29, 14, 6, 19, 1, 0, 0, 0], "xian3": [29, 14, 6, 19, 3, 0, 0, 0], "tong2": [25, 20, 19, 12, 2, 0, 0, 0], "ren4": [23, 10, 19, 4, 0, 0, 0, 0], "qian1": [22, 14, 6, 19, 1, 0, 0, 0], "dai4": [9, 6, 14, 4, 0, 0, 0, 0], "ling4": [17, 14, 19, 12, 4, 0, 0, 0], "lian2": [17, 14, 6, 19, 2, 0, 0, 0], "ling3": [17, 14, 19, 12, 3, 0, 0, 0], "si4": [24, 14, 4, 0, 0, 0, 0, 0], "sa1": [24, 6, 1, 0, 0, 0, 0, 0], "men": [18, 10, 19, 5, 0, 0, 0, 0], "fan3": [11, 6, 19, 3, 0, 0, 0, 0], "yang3": [30, 6, 19, 12, 3, 0, 0, 0], "ang2": [6, 19, 12, 2, 0, 0, 0, 0], "jian4": [15, 14, 6, 19, 4, 0, 0, 0], "mou2": [18, 20, 26, 2, 0, 0, 0, 0], "jia4": [15, 14, 6, 4, 0, 0, 0, 0], "jie": [15, 14, 10, 5, 0, 0, 0, 0], "lin4": [17, 14, 19, 4, 0, 0, 0, 0], "fen4": [11, 10, 19, 4, 0, 0, 0, 0], "bin1": [7, 14, 19, 1, 0, 0, 0, 0], "fang3": [11, 6, 19, 12, 3, 0, 0, 0], "pang2": [21, 6, 19, 12, 2, 0, 0, 0], "kang3": [16, 6, 19, 12, 3, 0, 0, 0], "ji4": [15, 14, 4, 0, 0, 0, 0, 0], "fu2": [11, 26, 2, 0, 0, 0, 0, 0], "xiu1": [29, 14, 26, 1, 0, 0, 0, 0], "xu4": [29, 26, 4, 0, 0, 0, 0, 0], "yin2": [30, 14, 19, 2, 0, 0, 0, 0], "you1": [30, 20, 26, 1, 0, 0, 0, 0], "you2": [30, 20, 26, 2, 0, 0, 0, 0], "huo3": [13, 26, 20, 3, 0, 0, 0, 0], "huo": [13, 26, 20, 5, 0, 0, 0, 0], "hui4": [13, 26, 14, 4, 0, 0, 0, 0], "kuai4": [16, 26, 6, 14, 4, 0, 0, 0], "san3": [24, 6, 19, 3, 0, 0, 0, 0], "wei3": [28, 10, 14, 3, 0, 0, 0, 0], "chuan2": [8, 13, 26, 6, 19, 2, 0, 0], "zhuan4": [31, 13, 26, 6, 19, 4, 0, 0], "ya2": [30, 6, 2, 0, 0, 0, 0, 0], "shang1": [24, 13, 6, 19, 12, 1, 0, 0], "ba4": [7, 6, 4, 0, 0, 0, 0, 0], "bai3": [7, 6, 14, 3, 0, 0, 0, 0], "gu1": [12, 26, 1, 0, 0, 0, 0, 0], "gu4": [12, 26, 4, 0, 0, 0, 0, 0], "pan4": [21, 6, 19, 4, 0, 0, 0, 0], "shen1": [24, 13, 10, 19, 1, 0, 0, 0], "ci4": [8, 14, 4, 0, 0, 0, 0, 0], "ga1": [12, 6, 1, 0, 0, 0, 0, 0], "jia1": [15, 14, 6, 1, 0, 0, 0, 0], "qie2": [22, 14, 10, 2, 0, 0, 0, 0], "dian4": [9, 14, 6, 19, 4, 0, 0, 0], "tian2": [25, 14, 6, 19, 2, 0, 0, 0], "dan4": [9, 6, 19, 4, 0, 0, 0, 0], "tan3": [25, 6, 19, 3, 0, 0, 0, 0], "yan4": [30, 6, 19, 4, 0, 0, 0, 0], "di1": [9, 14, 1, 0, 0, 0, 0, 0], "zuo3": [31, 26, 20, 3, 0, 0, 0, 0], "you4": [30, 20, 26, 4, 0, 0, 0, 0], "ti3": [25, 14, 3, 0, 0, 0, 0, 0], "ben4": [7, 10, 19, 4, 0, 0, 0, 0], "cui4": [8, 26, 14, 4, 0, 0, 0, 0], "ti1": [25, 14, 1, 0, 0, 0, 0, 0], "zhan4": [31, 13, 6, 19, 4, 0, 0, 0], "chan1": [8, 13, 6, 19, 1, 0, 0, 0], "dian1": [9, 14, 6, 19, 1, 0, 0, 0], "he2": [13, 10, 2, 0, 0, 0, 0, 0], "he4": [13, 10, 4, 0, 0, 0, 0, 0], "tuo1": [25, 26, 20, 1, 0, 0, 0, 0], "tuo4": [25, 26, 20, 4, 0, 0, 0, 0], "she2": [24, 13, 10, 2, 0, 0, 0, 0], "tu2": [25, 26, 2, 0, 0, 0, 0, 0], "xu2": [29, 26, 2, 0, 0, 0, 0, 0], "die2": [9, 14, 10, 2, 0, 0, 0, 0], "bi4": [7, 14, 4, 0, 0, 0, 0, 0], "zuo1": [31, 26, 20, 1, 0, 0, 0, 0], "zuo2": [31, 26, 20, 2, 0, 0, 0, 0], "gou1": [12, 20, 26, 1, 0, 0, 0, 0], "kou4": [16, 20, 26, 4, 0, 0, 0, 0], "ning4": [19, 14, 19, 12, 4, 0, 0, 0], "ni3": [19, 14, 3, 0, 0, 0, 0, 0], "qu2": [22, 26, 2, 0, 0, 0, 0, 0], "yong1": [30, 20, 19, 12, 1, 0, 0, 0], "yong4": [30, 20, 19, 12, 4, 0, 0, 0], "wa3": [28, 6, 3, 0, 0, 0, 0, 0], "pei4": [21, 10, 14, 4, 0, 0, 0, 0], "lao3": [17, 6, 20, 3, 0, 0, 0, 0], "liao2": [17, 14, 6, 20, 2, 0, 0, 0], "yang2": [30, 6, 19, 12, 2, 0, 0, 0], "tiao1": [25, 14, 6, 20, 1, 0, 0, 0], "tiao2": [25, 14, 6, 20, 2, 0, 0, 0], "tiao4": [25, 14, 6, 20, 4, 0, 0, 0], "diao3": [9, 14, 6, 20, 3, 0, 0, 0], "yao2": [30, 6, 20, 2, 0, 0, 0, 0], "dao4": [9, 6, 20, 4, 0, 0, 0, 0], "zhao4": [31, 13, 6, 20, 4, 0, 0, 0], "jiao3": [15, 14, 6, 20, 3, 0, 0, 0], "xiao2": [29, 14, 6, 20, 2, 0, 0, 0], "shi3": [24, 13, 14, 3, 0, 0, 0, 0], "kan3": [16, 6, 19, 3, 0, 0, 0, 0], "zhi2": [31, 13, 14, 2, 0, 0, 0, 0], "lai2": [17, 6, 14, 2, 0, 0, 0, 0], "lai4": [17, 6, 14, 4, 0, 0, 0, 0], "chi3": [8, 13, 14, 3, 0, 0, 0, 0], "lie4": [17, 14, 10, 4, 0, 0, 0, 0], "zhou1": [31, 13, 20, 26, 1, 0, 0, 0], "lun4": [17, 26, 19, 4, 0, 0, 0, 0], "dong4": [9, 20, 19, 12, 4, 0, 0, 0], "tong1": [25, 20, 19, 12, 1, 0, 0, 0], "tong3": [25, 20, 19, 12, 3, 0, 0, 0], "gong1": [12, 20, 19, 12, 1, 0, 0, 0], "gong4": [12, 20, 19, 12, 4, 0, 0, 0], "xia2": [29, 14, 6, 2, 0, 0, 0, 0], "lv3": [17, 27, 3, 0, 0, 0, 0, 0], "zhen1": [31, 13, 10, 19, 1, 0, 0, 0], "ce4": [8, 10, 4, 0, 0, 0, 0, 0], "zhai1": [31, 13, 6, 14, 1, 0, 0, 0], "nong2": [19, 20, 19, 12, 2, 0, 0, 0], "hou2": [13, 20, 26, 2, 0, 0, 0, 0], "hou4": [13, 20, 26, 4, 0, 0, 0, 0], "qin3": [22, 14, 19, 3, 0, 0, 0, 0], "ju2": [15, 26, 2, 0, 0, 0, 0, 0], "bian4": [7, 14, 6, 19, 4, 0, 0, 0], "pian2": [21, 14, 6, 19, 2, 0, 0, 0], "bian1": [7, 14, 6, 19, 1, 0, 0, 0], "xi4": [29, 14, 4, 0, 0, 0, 0, 0], "cu4": [8, 26, 4, 0, 0, 0, 0, 0], "chuo4": [8, 13, 26, 20, 4, 0, 0, 0], "e2": [10, 2, 0, 0, 0, 0, 0, 0], "jun4": [15, 26, 19, 4, 0, 0, 0, 0], "shun4": [24, 13, 26, 19, 4, 0, 0, 0], "dun1": [9, 26, 19, 1, 0, 0, 0, 0], "zu3": [31, 26, 3, 0, 0, 0, 0, 0], "qiao4": [22, 14, 6, 20, 4, 0, 0, 0], "xiao4": [29, 14, 6, 20, 4, 0, 0, 0], "xiao1": [29, 14, 6, 20, 1, 0, 0, 0], "yong3": [30, 20, 19, 12, 3, 0, 0, 0], "su2": [24, 26, 2, 0, 0, 0, 0, 0], "li3": [17, 14, 3, 0, 0, 0, 0, 0], "bao3": [7, 6, 20, 3, 0, 0, 0, 0], "shu4": [24, 13, 26, 4, 0, 0, 0, 0], "xin4": [29, 14, 19, 4, 0, 0, 0, 0], "yan3": [30, 6, 19, 3, 0, 0, 0, 0], "lia3": [17, 14, 6, 3, 0, 0, 0, 0], "jian3": [15, 14, 6, 19, 3, 0, 0, 0], "fu3": [11, 26, 3, 0, 0, 0, 0, 0], "ju4": [15, 26, 4, 0, 0, 0, 0, 0], "pai2": [21, 6, 14, 2, 0, 0, 0, 0], "feng4": [11, 10, 19, 12, 4, 0, 0, 0], "beng3": [7, 10, 19, 12, 3, 0, 0, 0], "an3": [6, 19, 3, 0, 0, 0, 0, 0], "bi3": [7, 14, 3, 0, 0, 0, 0, 0], "bei1": [7, 10, 14, 1, 0, 0, 0, 0], "pi4": [21, 14, 4, 0, 0, 0, 0, 0], "chuang4": [8, 13, 26, 6, 19, 12, 4, 0], "guan1": [12, 26, 6, 19, 1, 0, 0, 0], "bei4": [7, 10, 14, 4, 0, 0, 0, 0], "pei2": [21, 10, 14, 2, 0, 0, 0, 0], "men4": [18, 10, 19, 4, 0, 0, 0, 0], "men2": [18, 10, 19, 2, 0, 0, 0, 0], "dao3": [9, 6, 20, 3, 0, 0, 0, 0], "jue4": [15, 26, 10, 4, 0, 0, 0, 0], "xing4": [29, 14, 19, 12, 4, 0, 0, 0], "tang3": [25, 6, 19, 12, 3, 0, 0, 0], "chang2": [8, 13, 6, 19, 12, 2, 0, 0], "ti4": [25, 14, 4, 0, 0, 0, 0, 0], "diao4": [9, 14, 6, 20, 4, 0, 0, 0], "chang4": [8, 13, 6, 19, 12, 4, 0, 0], "chang1": [8, 13, 6, 19, 12, 1, 0, 0], "juan4": [15, 26, 6, 19, 4, 0, 0, 0], "qian4": [22, 14, 6, 19, 4, 0, 0, 0], "ni2": [19, 14, 2, 0, 0, 0, 0, 0], "ni4": [19, 14, 4, 0, 0, 0, 0, 0], "zhuo1": [31, 13, 26, 20, 1, 0, 0, 0], "wo1": [28, 20, 1, 0, 0, 0, 0, 0], "wei1": [28, 10, 14, 1, 0, 0, 0, 0], "wo3": [28, 20, 3, 0, 0, 0, 0, 0], "zhai4": [31, 13, 6, 14, 4, 0, 0, 0], "qing1": [22, 14, 19, 12, 1, 0, 0, 0], "jia3": [15, 14, 6, 3, 0, 0, 0, 0], "ge2": [12, 10, 2, 0, 0, 0, 0, 0], "jie2": [15, 14, 10, 2, 0, 0, 0, 0], "ruo4": [23, 26, 20, 4, 0, 0, 0, 0], "re4": [23, 10, 4, 0, 0, 0, 0, 0], "pian1": [21, 14, 6, 19, 1, 0, 0, 0], "xie2": [29, 14, 10, 2, 0, 0, 0, 0], "ou3": [20, 26, 3, 0, 0, 0, 0, 0], "tou1": [25, 20, 26, 1, 0, 0, 0, 0], "lou2": [17, 20, 26, 2, 0, 0, 0, 0], "gui4": [12, 26, 14, 4, 0, 0, 0, 0], "kui3": [16, 26, 14, 3, 0, 0, 0, 0], "beng1": [7, 10, 19, 12, 1, 0, 0, 0], "peng2": [21, 10, 19, 12, 2, 0, 0, 0], "xiang4": [29, 14, 6, 19, 12, 4, 0, 0], "dai3": [9, 6, 14, 3, 0, 0, 0, 0], "chu3": [8, 13, 26, 3, 0, 0, 0, 0], "nuo2": [19, 26, 20, 2, 0, 0, 0, 0], "cui1": [8, 26, 14, 1, 0, 0, 0, 0], "chong1": [8, 13, 20, 19, 12, 1, 0, 0], "ao4": [6, 20, 4, 0, 0, 0, 0, 0], "ao2": [6, 20, 2, 0, 0, 0, 0, 0], "sha3": [24, 13, 6, 3, 0, 0, 0, 0], "qing3": [22, 14, 19, 12, 3, 0, 0, 0], "xi1": [29, 14, 1, 0, 0, 0, 0, 0], "seng1": [24, 10, 19, 12, 1, 0, 0, 0], "ceng2": [8, 10, 19, 12, 2, 0, 0, 0], "zen4": [31, 10, 19, 4, 0, 0, 0, 0], "zhuang4": [31, 13, 26, 6, 19, 12, 4, 0], "chong4": [8, 13, 20, 19, 12, 4, 0, 0], "jiang1": [15, 14, 6, 19, 12, 1, 0, 0], "qia3": [22, 14, 6, 3, 0, 0, 0, 0], "shan4": [24, 13, 6, 19, 4, 0, 0, 0], "ru2": [23, 26, 2, 0, 0, 0, 0, 0], "chai2": [8, 13, 6, 14, 2, 0, 0, 0], "lei3": [17, 10, 14, 3, 0, 0, 0, 0], "lei2": [17, 10, 14, 2, 0, 0, 0, 0], "lei4": [17, 10, 14, 4, 0, 0, 0, 0], "chu2": [8, 13, 26, 2, 0, 0, 0, 0], "er2": [10, 23, 2, 0, 0, 0, 0, 0], "yun3": [30, 26, 19, 3, 0, 0, 0, 0], "xiong1": [29, 14, 20, 19, 12, 1, 0, 0], "kuang4": [16, 26, 6, 19, 12, 4, 0, 0], "guang1": [12, 26, 6, 19, 12, 1, 0, 0], "guang4": [12, 26, 6, 19, 12, 4, 0, 0], "ke4": [16, 10, 4, 0, 0, 0, 0, 0], "dui4": [9, 26, 14, 4, 0, 0, 0, 0], "mian3": [18, 14, 6, 19, 3, 0, 0, 0], "wen4": [28, 10, 19, 4, 0, 0, 0, 0], "wan3": [28, 6, 19, 3, 0, 0, 0, 0], "rui4": [23, 26, 14, 4, 0, 0, 0, 0], "duo2": [9, 26, 20, 2, 0, 0, 0, 0], "tu4": [25, 26, 4, 0, 0, 0, 0, 0], "dang3": [9, 6, 19, 12, 3, 0, 0, 0], "dou1": [9, 20, 26, 1, 0, 0, 0, 0], "ru4": [23, 26, 4, 0, 0, 0, 0, 0], "nei4": [19, 10, 14, 4, 0, 0, 0, 0], "quan2": [22, 26, 6, 19, 2, 0, 0, 0], "ba1": [7, 6, 1, 0, 0, 0, 0, 0], "ba2": [7, 6, 2, 0, 0, 0, 0, 0], "liu4": [17, 14, 26, 4, 0, 0, 0, 0], "lu4": [17, 26, 4, 0, 0, 0, 0, 0], "lan2": [17, 6, 19, 2, 0, 0, 0, 0], "gong3": [12, 20, 19, 12, 3, 0, 0, 0], "hong2": [13, 20, 19, 12, 2, 0, 0, 0], "tian1": [25, 14, 6, 19, 1, 0, 0, 0], "bing1": [7, 14, 19, 12, 1, 0, 0, 0], "dian3": [9, 14, 6, 19, 3, 0, 0, 0], "tian3": [25, 14, 6, 19, 3, 0, 0, 0], "ci2": [8, 14, 2, 0, 0, 0, 0, 0], "jian1": [15, 14, 6, 19, 1, 0, 0, 0], "shou4": [24, 13, 20, 26, 4, 0, 0, 0], "na4": [19, 6, 4, 0, 0, 0, 0, 0], "mao3": [18, 6, 20, 3, 0, 0, 0, 0], "ran3": [23, 6, 19, 3, 0, 0, 0, 0], "nan2": [19, 6, 19, 2, 0, 0, 0, 0], "zai4": [31, 6, 14, 4, 0, 0, 0, 0], "jiong3": [15, 14, 20, 19, 12, 3, 0, 0], "jiong1": [15, 14, 20, 19, 12, 1, 0, 0], "mao4": [18, 6, 20, 4, 0, 0, 0, 0], "rong3": [23, 20, 19, 12, 3, 0, 0, 0], "xie3": [29, 14, 10, 3, 0, 0, 0, 0], "jun1": [15, 26, 19, 1, 0, 0, 0, 0], "zhong3": [31, 13, 20, 19, 12, 3, 0, 0], "yuan1": [30, 26, 6, 19, 1, 0, 0, 0], "ming2": [18, 14, 19, 12, 2, 0, 0, 0], "mian2": [18, 14, 6, 19, 2, 0, 0, 0], "mian4": [18, 14, 6, 19, 4, 0, 0, 0], "mi4": [18, 14, 4, 0, 0, 0, 0, 0], "feng2": [11, 10, 19, 12, 2, 0, 0, 0], "ping2": [21, 14, 19, 12, 2, 0, 0, 0], "ning2": [19, 14, 19, 12, 2, 0, 0, 0], "leng3": [17, 10, 19, 12, 3, 0, 0, 0], "sheng3": [24, 13, 10, 19, 12, 3, 0, 0], "qia4": [22, 14, 6, 4, 0, 0, 0, 0], "cheng1": [8, 13, 10, 19, 12, 1, 0, 0], "zhun3": [31, 13, 26, 19, 3, 0, 0, 0], "song1": [24, 20, 19, 12, 1, 0, 0, 0], "diao1": [9, 14, 6, 20, 1, 0, 0, 0], "cou4": [8, 20, 26, 4, 0, 0, 0, 0], "lin3": [17, 14, 19, 3, 0, 0, 0, 0], "ji3": [15, 14, 3, 0, 0, 0, 0, 0], "fan2": [11, 6, 19, 2, 0, 0, 0, 0], "chu4": [8, 13, 26, 4, 0, 0, 0, 0], "zhi3": [31, 13, 14, 3, 0, 0, 0, 0], "kai3": [16, 6, 14, 3, 0, 0, 0, 0], "huang2": [13, 26, 6, 19, 12, 2, 0, 0], "deng4": [9, 10, 19, 12, 4, 0, 0, 0], "tu1": [25, 26, 1, 0, 0, 0, 0, 0], "ao1": [6, 20, 1, 0, 0, 0, 0, 0], "wa1": [28, 6, 1, 0, 0, 0, 0, 0], "chu1": [8, 13, 26, 1, 0, 0, 0, 0], "han2": [13, 6, 19, 2, 0, 0, 0, 0], "zao2": [31, 6, 20, 2, 0, 0, 0, 0], "dao1": [9, 6, 20, 1, 0, 0, 0, 0], "fen1": [11, 10, 19, 1, 0, 0, 0, 0], "fen2": [11, 10, 19, 2, 0, 0, 0, 0], "qie4": [22, 14, 10, 4, 0, 0, 0, 0], "qie1": [22, 14, 10, 1, 0, 0, 0, 0], "kan1": [16, 6, 19, 1, 0, 0, 0, 0], "wen3": [28, 10, 19, 3, 0, 0, 0, 0], "xing2": [29, 14, 19, 12, 2, 0, 0, 0], "hua4": [13, 26, 6, 4, 0, 0, 0, 0], "guo4": [12, 26, 20, 4, 0, 0, 0, 0], "guo3": [12, 26, 20, 3, 0, 0, 0, 0], "hua2": [13, 26, 6, 2, 0, 0, 0, 0], "huai": [13, 26, 6, 14, 5, 0, 0, 0], "liu2": [17, 14, 26, 2, 0, 0, 0, 0], "ze2": [31, 10, 2, 0, 0, 0, 0, 0], "chuang1": [8, 13, 26, 6, 19, 12, 1, 0], "shan1": [24, 13, 6, 19, 1, 0, 0, 0], "bie2": [7, 14, 10, 2, 0, 0, 0, 0], "pao2": [21, 6, 20, 2, 0, 0, 0, 0], "bao4": [7, 6, 20, 4, 0, 0, 0, 0], "bie4": [7, 14, 10, 4, 0, 0, 0, 0], "gua1": [12, 26, 6, 1, 0, 0, 0, 0], "shua1": [24, 13, 26, 6, 1, 0, 0, 0], "shua4": [24, 13, 26, 6, 4, 0, 0, 0], "xuan4": [29, 26, 6, 19, 4, 0, 0, 0], "sha1": [24, 13, 6, 1, 0, 0, 0, 0], "cha4": [8, 13, 6, 4, 0, 0, 0, 0], "ci1": [8, 14, 1, 0, 0, 0, 0, 0], "kei1": [16, 10, 14, 1, 0, 0, 0, 0], "duo4": [9, 26, 20, 4, 0, 0, 0, 0], "cuo4": [8, 26, 20, 4, 0, 0, 0, 0], "xue1": [29, 26, 10, 1, 0, 0, 0, 0], "shao4": [24, 13, 6, 20, 4, 0, 0, 0], "la2": [17, 6, 2, 0, 0, 0, 0, 0], "la4": [17, 6, 4, 0, 0, 0, 0, 0], "gua3": [12, 26, 6, 3, 0, 0, 0, 0], "pou1": [21, 20, 26, 1, 0, 0, 0, 0], "po3": [21, 20, 3, 0, 0, 0, 0, 0], "wan1": [28, 6, 19, 1, 0, 0, 0, 0], "bo1": [7, 20, 1, 0, 0, 0, 0, 0], "bao1": [7, 6, 20, 1, 0, 0, 0, 0], "ge1": [12, 10, 1, 0, 0, 0, 0, 0], "qiang1": [22, 14, 6, 19, 12, 1, 0, 0], "piao1": [21, 14, 6, 20, 1, 0, 0, 0], "piao4": [21, 14, 6, 20, 4, 0, 0, 0], "piao2": [21, 14, 6, 20, 2, 0, 0, 0], "biao3": [7, 14, 6, 20, 3, 0, 0, 0], "biao1": [7, 14, 6, 20, 1, 0, 0, 0], "chao1": [8, 13, 6, 20, 1, 0, 0, 0], "pi3": [21, 14, 3, 0, 0, 0, 0, 0], "tang1": [25, 6, 19, 12, 1, 0, 0, 0], "nu3": [19, 26, 3, 0, 0, 0, 0, 0], "lao2": [17, 6, 20, 2, 0, 0, 0, 0], "kai4": [16, 6, 14, 4, 0, 0, 0, 0], "xun1": [29, 26, 19, 1, 0, 0, 0, 0], "meng3": [18, 10, 19, 12, 3, 0, 0, 0], "lei1": [17, 10, 14, 1, 0, 0, 0, 0], "lei": [17, 10, 14, 5, 0, 0, 0, 0], "mao2": [18, 6, 20, 2, 0, 0, 0, 0], "lao4": [17, 6, 20, 4, 0, 0, 0, 0], "mu4": [18, 26, 4, 0, 0, 0, 0, 0], "qin2": [22, 14, 19, 2, 0, 0, 0, 0], "shao2": [24, 13, 6, 20, 2, 0, 0, 0], "shuo4": [24, 13, 26, 20, 4, 0, 0, 0], "zhuo2": [31, 13, 26, 20, 2, 0, 0, 0], "di4": [9, 14, 4, 0, 0, 0, 0, 0], "gou4": [12, 20, 26, 4, 0, 0, 0, 0], "yun4": [30, 26, 19, 4, 0, 0, 0, 0], "pin4": [21, 14, 19, 4, 0, 0, 0, 0], "hua1": [13, 26, 6, 1, 0, 0, 0, 0], "huo4": [13, 26, 20, 4, 0, 0, 0, 0], "bei3": [7, 10, 14, 3, 0, 0, 0, 0], "shi": [24, 13, 14, 5, 0, 0, 0, 0], "chi2": [8, 13, 14, 2, 0, 0, 0, 0], "za1": [31, 6, 1, 0, 0, 0, 0, 0], "jiang4": [15, 14, 6, 19, 12, 4, 0, 0], "kuang1": [16, 26, 6, 19, 12, 1, 0, 0], "wang1": [28, 6, 19, 12, 1, 0, 0, 0], "fei3": [11, 10, 14, 3, 0, 0, 0, 0], "fei1": [11, 10, 14, 1, 0, 0, 0, 0], "kui4": [16, 26, 14, 4, 0, 0, 0, 0], "qu1": [22, 26, 1, 0, 0, 0, 0, 0], "ou1": [20, 26, 1, 0, 0, 0, 0, 0], "bian3": [7, 14, 6, 19, 3, 0, 0, 0], "te4": [25, 10, 4, 0, 0, 0, 0, 0], "sa4": [24, 6, 4, 0, 0, 0, 0, 0], "sheng1": [24, 13, 10, 19, 12, 1, 0, 0], "pi2": [21, 14, 2, 0, 0, 0, 0, 0], "ban1": [7, 6, 19, 1, 0, 0, 0, 0], "zu2": [31, 26, 2, 0, 0, 0, 0, 0], "chan2": [8, 13, 6, 19, 2, 0, 0, 0], "mai4": [18, 6, 14, 4, 0, 0, 0, 0], "na1": [19, 6, 1, 0, 0, 0, 0, 0], "bo": [7, 20, 5, 0, 0, 0, 0, 0], "bu3": [7, 26, 3, 0, 0, 0, 0, 0], "pan2": [21, 6, 19, 2, 0, 0, 0, 0], "zhan1": [31, 13, 6, 19, 1, 0, 0, 0], "tie1": [25, 14, 10, 1, 0, 0, 0, 0], "ka3": [16, 6, 3, 0, 0, 0, 0, 0], "lu2": [17, 26, 2, 0, 0, 0, 0, 0], "lu3": [17, 26, 3, 0, 0, 0, 0, 0], "gua4": [12, 26, 6, 4, 0, 0, 0, 0], "wo4": [28, 20, 4, 0, 0, 0, 0, 0], "yin4": [30, 14, 19, 4, 0, 0, 0, 0], "que4": [22, 26, 10, 4, 0, 0, 0, 0], "luan3": [17, 26, 6, 19, 3, 0, 0, 0], "kun1": [16, 26, 19, 1, 0, 0, 0, 0], "juan3": [15, 26, 6, 19, 3, 0, 0, 0], "quan1": [22, 26, 6, 19, 1, 0, 0, 0], "chang3": [8, 13, 6, 19, 12, 3, 0, 0], "han3": [13, 6, 19, 3, 0, 0, 0, 0], "an1": [6, 19, 1, 0, 0, 0, 0, 0], "e3": [10, 3, 0, 0, 0, 0, 0, 0], "ting1": [25, 14, 19, 12, 1, 0, 0, 0], "si": [24, 14, 5, 0, 0, 0, 0, 0], "sha4": [24, 13, 6, 4, 0, 0, 0, 0], "jiu4": [15, 14, 26, 4, 0, 0, 0, 0], "yan1": [30, 6, 19, 1, 0, 0, 0, 0], "qu4": [22, 26, 4, 0, 0, 0, 0, 0], "xian4": [29, 14, 6, 19, 4, 0, 0, 0], "can1": [8, 6, 19, 1, 0, 0, 0, 0], "cen1": [8, 10, 19, 1, 0, 0, 0, 0], "can4": [8, 6, 19, 4, 0, 0, 0, 0], "cha1": [8, 13, 6, 1, 0, 0, 0, 0], "cha2": [8, 13, 6, 2, 0, 0, 0, 0], "cha3": [8, 13, 6, 3, 0, 0, 0, 0], "you3": [30, 20, 26, 3, 0, 0, 0, 0], "shuang1": [24, 13, 26, 6, 19, 12, 1, 0], "fan4": [11, 6, 19, 4, 0, 0, 0, 0], "shou1": [24, 13, 20, 26, 1, 0, 0, 0], "fa1": [11, 6, 1, 0, 0, 0, 0, 0], "fa4": [11, 6, 4, 0, 0, 0, 0, 0], "qu3": [22, 26, 3, 0, 0, 0, 0, 0], "sou3": [24, 20, 26, 3, 0, 0, 0, 0], "sou1": [24, 20, 26, 1, 0, 0, 0, 0], "kou3": [16, 20, 26, 3, 0, 0, 0, 0], "gu3": [12, 26, 3, 0, 0, 0, 0, 0], "ku1": [16, 26, 1, 0, 0, 0, 0, 0], "tao1": [25, 6, 20, 1, 0, 0, 0, 0], "dao2": [9, 6, 20, 2, 0, 0, 0, 0], "jiao4": [15, 14, 6, 20, 4, 0, 0, 0], "pa1": [21, 6, 1, 0, 0, 0, 0, 0], "ba": [7, 6, 5, 0, 0, 0, 0, 0], "ke3": [16, 10, 3, 0, 0, 0, 0, 0], "tai2": [25, 6, 14, 2, 0, 0, 0, 0], "tai1": [25, 6, 14, 1, 0, 0, 0, 0], "chi4": [8, 13, 14, 4, 0, 0, 0, 0], "hao4": [13, 6, 20, 4, 0, 0, 0, 0], "hao2": [13, 6, 20, 2, 0, 0, 0, 0], "tan4": [25, 6, 19, 4, 0, 0, 0, 0], "chi1": [8, 13, 14, 1, 0, 0, 0, 0], "cun4": [8, 26, 19, 4, 0, 0, 0, 0], "dou4": [9, 20, 26, 4, 0, 0, 0, 0], "ying1": [30, 14, 19, 12, 1, 0, 0, 0], "tong4": [25, 20, 19, 12, 4, 0, 0, 0], "ming4": [18, 14, 19, 12, 4, 0, 0, 0], "tu3": [25, 26, 3, 0, 0, 0, 0, 0], "zha1": [31, 13, 6, 1, 0, 0, 0, 0], "ha4": [13, 6, 4, 0, 0, 0, 0, 0], "a1": [6, 1, 0, 0, 0, 0, 0, 0], "ma2": [18, 6, 2, 0, 0, 0, 0, 0], "ma3": [18, 6, 3, 0, 0, 0, 0, 0], "tun1": [25, 26, 19, 1, 0, 0, 0, 0], "yin3": [30, 14, 19, 3, 0, 0, 0, 0], "fei4": [11, 10, 14, 4, 0, 0, 0, 0], "tun2": [25, 26, 19, 2, 0, 0, 0, 0], "tun3": [25, 26, 19, 3, 0, 0, 0, 0], "pen4": [21, 10, 19, 4, 0, 0, 0, 0], "han4": [13, 6, 19, 4, 0, 0, 0, 0], "keng1": [16, 10, 19, 12, 1, 0, 0, 0], "hang2": [13, 6, 19, 12, 2, 0, 0, 0], "hang4": [13, 6, 19, 12, 4, 0, 0, 0], "shun3": [24, 13, 26, 19, 3, 0, 0, 0], "chao3": [8, 13, 6, 20, 3, 0, 0, 0], "miao3": [18, 14, 6, 20, 3, 0, 0, 0], "chao4": [8, 13, 6, 20, 4, 0, 0, 0], "chui1": [8, 13, 26, 14, 1, 0, 0, 0], "chui4": [8, 13, 26, 14, 4, 0, 0, 0], "hou3": [13, 20, 26, 3, 0, 0, 0, 0], "hong1": [13, 20, 19, 12, 1, 0, 0, 0], "ya": [30, 6, 5, 0, 0, 0, 0, 0], "xia1": [29, 14, 6, 1, 0, 0, 0, 0], "e": [10, 5, 0, 0, 0, 0, 0, 0], "dai1": [9, 6, 14, 1, 0, 0, 0, 0], "ai2": [6, 14, 2, 0, 0, 0, 0, 0], "kuang2": [16, 26, 6, 19, 12, 2, 0, 0], "cheng3": [8, 13, 10, 19, 12, 3, 0, 0], "gao4": [12, 6, 20, 4, 0, 0, 0, 0], "ne4": [19, 10, 4, 0, 0, 0, 0, 0], "na": [19, 6, 5, 0, 0, 0, 0, 0], "nuo4": [19, 26, 20, 4, 0, 0, 0, 0], "ne": [19, 10, 5, 0, 0, 0, 0, 0], "ou4": [20, 26, 4, 0, 0, 0, 0, 0], "bei": [7, 10, 14, 5, 0, 0, 0, 0], "bai4": [7, 6, 14, 4, 0, 0, 0, 0], "qiang4": [22, 14, 6, 19, 12, 4, 0, 0], "ni1": [19, 14, 1, 0, 0, 0, 0, 0], "mei4": [18, 10, 14, 4, 0, 0, 0, 0], "he1": [13, 10, 1, 0, 0, 0, 0, 0], "ha1": [13, 6, 1, 0, 0, 0, 0, 0], "a": [6, 5, 0, 0, 0, 0, 0, 0], "ke1": [16, 10, 1, 0, 0, 0, 0, 0], "huo1": [13, 26, 20, 1, 0, 0, 0, 0], "a2": [6, 2, 0, 0, 0, 0, 0, 0], "a4": [6, 4, 0, 0, 0, 0, 0, 0], "pei1": [21, 10, 14, 1, 0, 0, 0, 0], "zui3": [31, 26, 14, 3, 0, 0, 0, 0], "xian2": [29, 14, 6, 19, 2, 0, 0, 0], "duo1": [9, 26, 20, 1, 0, 0, 0, 0], "za3": [31, 6, 3, 0, 0, 0, 0, 0], "huo2": [13, 26, 20, 2, 0, 0, 0, 0], "hu2": [13, 26, 2, 0, 0, 0, 0, 0], "gao1": [12, 6, 20, 1, 0, 0, 0, 0], "zhou4": [31, 13, 20, 26, 4, 0, 0, 0], "ka1": [16, 6, 1, 0, 0, 0, 0, 0], "nong4": [19, 20, 19, 12, 4, 0, 0, 0], "gu": [12, 26, 5, 0, 0, 0, 0, 0], "zuo": [31, 26, 20, 5, 0, 0, 0, 0], "long2": [17, 20, 19, 12, 2, 0, 0, 0], "lie3": [17, 14, 10, 3, 0, 0, 0, 0], "lie1": [17, 14, 10, 1, 0, 0, 0, 0], "lie2": [17, 14, 10, 2, 0, 0, 0, 0], "lie": [17, 14, 10, 5, 0, 0, 0, 0], "mie": [18, 14, 10, 5, 0, 0, 0, 0], "mi1": [18, 14, 1, 0, 0, 0, 0, 0], "mi3": [18, 14, 3, 0, 0, 0, 0, 0], "yao3": [30, 6, 20, 3, 0, 0, 0, 0], "luo4": [17, 26, 20, 4, 0, 0, 0, 0], "lo": [17, 20, 5, 0, 0, 0, 0, 0], "zan2": [31, 6, 19, 2, 0, 0, 0, 0], "za2": [31, 6, 2, 0, 0, 0, 0, 0], "zan": [31, 6, 19, 5, 0, 0, 0, 0], "hai1": [13, 6, 14, 1, 0, 0, 0, 0], "hai2": [13, 6, 14, 2, 0, 0, 0, 0], "ke2": [16, 10, 2, 0, 0, 0, 0, 0], "gai1": [12, 6, 14, 1, 0, 0, 0, 0], "xu3": [29, 26, 3, 0, 0, 0, 0, 0], "ai1": [6, 14, 1, 0, 0, 0, 0, 0], "pin3": [21, 14, 19, 3, 0, 0, 0, 0], "shen3": [24, 13, 10, 19, 3, 0, 0, 0], "hong4": [13, 20, 19, 12, 4, 0, 0, 0], "hong3": [13, 20, 19, 12, 3, 0, 0, 0], "die3": [9, 14, 10, 3, 0, 0, 0, 0], "wa": [28, 6, 5, 0, 0, 0, 0, 0], "wa2": [28, 6, 2, 0, 0, 0, 0, 0], "ha3": [13, 6, 3, 0, 0, 0, 0, 0], "ta4": [25, 6, 4, 0, 0, 0, 0, 0], "zai1": [31, 6, 14, 1, 0, 0, 0, 0], "die4": [9, 14, 10, 4, 0, 0, 0, 0], "pai4": [21, 6, 14, 4, 0, 0, 0, 0], "gen2": [12, 10, 19, 2, 0, 0, 0, 0], "hen3": [13, 10, 19, 3, 0, 0, 0, 0], "n4": [19, 4, 0, 0, 0, 0, 0, 0], "ya3": [30, 6, 3, 0, 0, 0, 0, 0], "da2": [9, 6, 2, 0, 0, 0, 0, 0], "yo1": [30, 20, 1, 0, 0, 0, 0, 0], "yo": [30, 20, 5, 0, 0, 0, 0, 0], "o2": [20, 2, 0, 0, 0, 0, 0, 0], "o4": [20, 4, 0, 0, 0, 0, 0, 0], "sao1": [24, 6, 20, 1, 0, 0, 0, 0], "sao4": [24, 6, 20, 4, 0, 0, 0, 0], "li1": [17, 14, 1, 0, 0, 0, 0, 0], "li": [17, 14, 5, 0, 0, 0, 0, 0], "mai2": [18, 6, 14, 2, 0, 0, 0, 0], "na3": [19, 6, 3, 0, 0, 0, 0, 0], "ne2": [19, 10, 2, 0, 0, 0, 0, 0], "nei3": [19, 10, 14, 3, 0, 0, 0, 0], "zhe2": [31, 13, 10, 2, 0, 0, 0, 0], "bu1": [7, 26, 1, 0, 0, 0, 0, 0], "hng": [13, 19, 12, 5, 0, 0, 0, 0], "geng3": [12, 10, 19, 12, 3, 0, 0, 0], "ying3": [30, 14, 19, 12, 3, 0, 0, 0], "ying4": [30, 14, 19, 12, 4, 0, 0, 0], "ng2": [19, 12, 2, 0, 0, 0, 0, 0], "n2": [19, 2, 0, 0, 0, 0, 0, 0], "chun2": [8, 13, 26, 19, 2, 0, 0, 0], "zhen4": [31, 13, 10, 19, 4, 0, 0, 0], "tang2": [25, 6, 19, 12, 2, 0, 0, 0], "m2": [18, 2, 0, 0, 0, 0, 0, 0], "huan4": [13, 26, 6, 19, 4, 0, 0, 0], "hu3": [13, 26, 3, 0, 0, 0, 0, 0], "guo2": [12, 26, 20, 2, 0, 0, 0, 0], "shu2": [24, 13, 26, 2, 0, 0, 0, 0], "nian4": [19, 14, 6, 19, 4, 0, 0, 0], "ken3": [16, 10, 19, 3, 0, 0, 0, 0], "len4": [17, 10, 19, 4, 0, 0, 0, 0], "a3": [6, 3, 0, 0, 0, 0, 0, 0], "tao2": [25, 6, 20, 2, 0, 0, 0, 0], "chuai4": [8, 13, 26, 6, 14, 4, 0, 0], "sha2": [24, 13, 6, 2, 0, 0, 0, 0], "la": [17, 6, 5, 0, 0, 0, 0, 0], "la1": [17, 6, 1, 0, 0, 0, 0, 0], "zhe3": [31, 13, 10, 3, 0, 0, 0, 0], "se4": [24, 10, 4, 0, 0, 0, 0, 0], "luo1": [17, 26, 20, 1, 0, 0, 0, 0], "ding4": [9, 14, 19, 12, 4, 0, 0, 0], "lang1": [17, 6, 19, 12, 1, 0, 0, 0], "ti2": [25, 14, 2, 0, 0, 0, 0, 0], "ke": [16, 10, 5, 0, 0, 0, 0, 0], "nan3": [19, 6, 19, 3, 0, 0, 0, 0], "la3": [17, 6, 3, 0, 0, 0, 0, 0], "kan4": [16, 6, 19, 4, 0, 0, 0, 0], "zha2": [31, 13, 6, 2, 0, 0, 0, 0], "wai1": [28, 6, 14, 1, 0, 0, 0, 0], "re3": [23, 10, 3, 0, 0, 0, 0, 0], "o1": [20, 1, 0, 0, 0, 0, 0, 0], "o": [20, 5, 0, 0, 0, 0, 0, 0], "chuan3": [8, 13, 26, 6, 19, 3, 0, 0], "xi3": [29, 14, 3, 0, 0, 0, 0, 0], "huai4": [13, 26, 6, 14, 4, 0, 0, 0], "xuan3": [29, 26, 6, 19, 3, 0, 0, 0], "tan2": [25, 6, 19, 2, 0, 0, 0, 0], "sun1": [24, 26, 19, 1, 0, 0, 0, 0], "zha": [31, 13, 6, 5, 0, 0, 0, 0], "miao1": [18, 14, 6, 20, 1, 0, 0, 0], "ying2": [30, 14, 19, 12, 2, 0, 0, 0], "pen1": [21, 10, 19, 1, 0, 0, 0, 0], "kui2": [16, 26, 14, 2, 0, 0, 0, 0], "lou": [17, 20, 26, 5, 0, 0, 0, 0], "xiu4": [29, 14, 26, 4, 0, 0, 0, 0], "ma4": [18, 6, 4, 0, 0, 0, 0, 0], "da1": [9, 6, 1, 0, 0, 0, 0, 0], "da": [9, 6, 5, 0, 0, 0, 0, 0], "sang3": [24, 6, 19, 12, 3, 0, 0, 0], "chen1": [8, 13, 10, 19, 1, 0, 0, 0], "su4": [24, 26, 4, 0, 0, 0, 0, 0], "sou4": [24, 20, 26, 4, 0, 0, 0, 0], "jue1": [15, 26, 10, 1, 0, 0, 0, 0], "weng1": [28, 10, 19, 12, 1, 0, 0, 0], "weng3": [28, 10, 19, 12, 3, 0, 0, 0], "suo": [24, 26, 20, 5, 0, 0, 0, 0], "hei1": [13, 10, 14, 1, 0, 0, 0, 0], "ng3": [19, 12, 3, 0, 0, 0, 0, 0], "n3": [19, 3, 0, 0, 0, 0, 0, 0], "ng4": [19, 12, 4, 0, 0, 0, 0, 0], "die1": [9, 14, 10, 1, 0, 0, 0, 0], "dia3": [9, 14, 6, 3, 0, 0, 0, 0], "di2": [9, 14, 2, 0, 0, 0, 0, 0], "cao2": [8, 6, 20, 2, 0, 0, 0, 0], "lou3": [17, 20, 26, 3, 0, 0, 0, 0], "ga2": [12, 6, 2, 0, 0, 0, 0, 0], "ga3": [12, 6, 3, 0, 0, 0, 0, 0], "ou": [20, 26, 5, 0, 0, 0, 0, 0], "shi1": [24, 13, 14, 1, 0, 0, 0, 0], "de1": [9, 10, 1, 0, 0, 0, 0, 0], "de2": [9, 10, 2, 0, 0, 0, 0, 0], "dei1": [9, 10, 14, 1, 0, 0, 0, 0], "le1": [17, 10, 1, 0, 0, 0, 0, 0], "du1": [9, 26, 1, 0, 0, 0, 0, 0], "chao2": [8, 13, 6, 20, 2, 0, 0, 0], "zhao1": [31, 13, 6, 20, 1, 0, 0, 0], "m1": [18, 1, 0, 0, 0, 0, 0, 0], "ceng1": [8, 10, 19, 12, 1, 0, 0, 0], "ye1": [30, 10, 1, 0, 0, 0, 0, 0], "deng1": [9, 10, 19, 12, 1, 0, 0, 0], "lu1": [17, 26, 1, 0, 0, 0, 0, 0], "zao4": [31, 6, 20, 4, 0, 0, 0, 0], "xue2": [29, 26, 10, 2, 0, 0, 0, 0], "dang1": [9, 6, 19, 12, 1, 0, 0, 0], "sai1": [24, 6, 14, 1, 0, 0, 0, 0], "ca1": [8, 6, 1, 0, 0, 0, 0, 0], "rang3": [23, 6, 19, 12, 3, 0, 0, 0], "rang1": [23, 6, 19, 12, 1, 0, 0, 0], "jiao2": [15, 14, 6, 20, 2, 0, 0, 0], "luo2": [17, 26, 20, 2, 0, 0, 0, 0], "luo": [17, 26, 20, 5, 0, 0, 0, 0], "nang2": [19, 6, 19, 12, 2, 0, 0, 0], "nang1": [19, 6, 19, 12, 1, 0, 0, 0], "nang": [19, 6, 19, 12, 5, 0, 0, 0], "nan1": [19, 6, 19, 1, 0, 0, 0, 0], "hui2": [13, 26, 14, 2, 0, 0, 0, 0], "yin1": [30, 14, 19, 1, 0, 0, 0, 0], "tuan2": [25, 26, 6, 19, 2, 0, 0, 0], "dun4": [9, 26, 19, 4, 0, 0, 0, 0], "kun4": [16, 26, 19, 4, 0, 0, 0, 0], "pu3": [21, 26, 3, 0, 0, 0, 0, 0], "juan1": [15, 26, 6, 19, 1, 0, 0, 0], "huan2": [13, 26, 6, 19, 2, 0, 0, 0], "du4": [9, 26, 4, 0, 0, 0, 0, 0], "de": [9, 10, 5, 0, 0, 0, 0, 0], "quan3": [22, 26, 6, 19, 3, 0, 0, 0], "huai2": [13, 26, 6, 14, 2, 0, 0, 0], "ban3": [7, 6, 19, 3, 0, 0, 0, 0], "fang1": [11, 6, 19, 12, 1, 0, 0, 0], "fang2": [11, 6, 19, 12, 2, 0, 0, 0], "tan1": [25, 6, 19, 1, 0, 0, 0, 0], "yue2": [30, 26, 10, 2, 0, 0, 0, 0], "zhui4": [31, 13, 26, 14, 4, 0, 0, 0], "po1": [21, 20, 1, 0, 0, 0, 0, 0], "chui2": [8, 13, 26, 14, 2, 0, 0, 0], "long3": [17, 20, 19, 12, 3, 0, 0, 0], "duo3": [9, 26, 20, 3, 0, 0, 0, 0], "ken4": [16, 10, 19, 4, 0, 0, 0, 0], "kua3": [16, 26, 6, 3, 0, 0, 0, 0], "man2": [18, 6, 19, 2, 0, 0, 0, 0], "pou3": [21, 20, 26, 3, 0, 0, 0, 0], "dui1": [9, 26, 14, 1, 0, 0, 0, 0], "zui1": [31, 26, 14, 1, 0, 0, 0, 0], "hui1": [13, 26, 14, 1, 0, 0, 0, 0], "pu4": [21, 26, 4, 0, 0, 0, 0, 0], "di3": [9, 14, 3, 0, 0, 0, 0, 0], "chen3": [8, 13, 10, 19, 3, 0, 0, 0], "dang4": [9, 6, 19, 12, 4, 0, 0, 0], "du3": [9, 26, 3, 0, 0, 0, 0, 0], "ta3": [25, 6, 3, 0, 0, 0, 0, 0], "sai4": [24, 6, 14, 4, 0, 0, 0, 0], "chen2": [8, 13, 10, 19, 2, 0, 0, 0], "wen1": [28, 10, 19, 1, 0, 0, 0, 0], "zeng1": [31, 10, 19, 12, 1, 0, 0, 0], "qiang2": [22, 14, 6, 19, 12, 2, 0, 0], "zeng4": [31, 10, 19, 12, 4, 0, 0, 0], "weng4": [28, 10, 19, 12, 4, 0, 0, 0], "lv4": [17, 27, 4, 0, 0, 0, 0, 0], "zhuang1": [31, 13, 26, 6, 19, 12, 1, 0], "wai4": [28, 6, 14, 4, 0, 0, 0, 0], "meng4": [18, 10, 19, 12, 4, 0, 0, 0], "meng2": [18, 10, 19, 12, 2, 0, 0, 0], "da4": [9, 6, 4, 0, 0, 0, 0, 0], "tai4": [25, 6, 14, 4, 0, 0, 0, 0], "yang1": [30, 6, 19, 12, 1, 0, 0, 0], "hang1": [13, 6, 19, 12, 1, 0, 0, 0], "tou2": [25, 20, 26, 2, 0, 0, 0, 0], "tou": [25, 20, 26, 5, 0, 0, 0, 0], "kua1": [16, 26, 6, 1, 0, 0, 0, 0], "kua4": [16, 26, 6, 4, 0, 0, 0, 0], "jia2": [15, 14, 6, 2, 0, 0, 0, 0], "nai4": [19, 6, 14, 4, 0, 0, 0, 0], "zou4": [31, 20, 26, 4, 0, 0, 0, 0], "ben1": [7, 10, 19, 1, 0, 0, 0, 0], "jiang3": [15, 14, 6, 19, 12, 3, 0, 0], "tao4": [25, 6, 20, 4, 0, 0, 0, 0], "tao3": [25, 6, 20, 3, 0, 0, 0, 0], "zang4": [31, 6, 19, 12, 4, 0, 0, 0], "zhuang3": [31, 13, 26, 6, 19, 12, 3, 0], "zheng4": [31, 13, 10, 19, 12, 4, 0, 0], "zun1": [31, 26, 19, 1, 0, 0, 0, 0], "she1": [24, 13, 10, 1, 0, 0, 0, 0], "nv3": [19, 27, 3, 0, 0, 0, 0, 0], "nv4": [19, 27, 4, 0, 0, 0, 0, 0], "jie3": [15, 14, 10, 3, 0, 0, 0, 0], "hao3": [13, 6, 20, 3, 0, 0, 0, 0], "wang4": [28, 6, 19, 12, 4, 0, 0, 0], "ma1": [18, 6, 1, 0, 0, 0, 0, 0], "miao4": [18, 14, 6, 20, 4, 0, 0, 0], "niu1": [19, 14, 26, 1, 0, 0, 0, 0], "tuo3": [25, 26, 20, 3, 0, 0, 0, 0], "pan1": [21, 6, 19, 1, 0, 0, 0, 0], "zu1": [31, 26, 1, 0, 0, 0, 0, 0], "yao4": [30, 6, 20, 4, 0, 0, 0, 0], "zhen3": [31, 13, 10, 19, 3, 0, 0, 0], "rao2": [23, 6, 20, 2, 0, 0, 0, 0], "rao3": [23, 6, 20, 3, 0, 0, 0, 0], "suo3": [24, 26, 20, 3, 0, 0, 0, 0], "niang2": [19, 14, 6, 19, 12, 2, 0, 0], "e1": [10, 1, 0, 0, 0, 0, 0, 0], "lv2": [17, 27, 2, 0, 0, 0, 0, 0], "po2": [21, 20, 2, 0, 0, 0, 0, 0], "hun1": [13, 26, 19, 1, 0, 0, 0, 0], "lan3": [17, 6, 19, 3, 0, 0, 0, 0], "mei2": [18, 10, 14, 2, 0, 0, 0, 0], "yuan4": [30, 26, 6, 19, 4, 0, 0, 0], "bi1": [7, 14, 1, 0, 0, 0, 0, 0], "sao3": [24, 6, 20, 3, 0, 0, 0, 0], "bao2": [7, 6, 20, 2, 0, 0, 0, 0], "pin2": [21, 14, 19, 2, 0, 0, 0, 0], "man1": [18, 6, 19, 1, 0, 0, 0, 0], "man4": [18, 6, 19, 4, 0, 0, 0, 0], "nen4": [19, 10, 19, 4, 0, 0, 0, 0], "qiong2": [22, 14, 20, 19, 12, 2, 0, 0], "xuan2": [29, 26, 6, 19, 2, 0, 0, 0], "rang2": [23, 6, 19, 12, 2, 0, 0, 0], "zi": [31, 14, 5, 0, 0, 0, 0, 0], "kong3": [16, 20, 19, 12, 3, 0, 0, 0], "cun2": [8, 26, 19, 2, 0, 0, 0, 0], "luan2": [17, 26, 6, 19, 2, 0, 0, 0], "xun4": [29, 26, 19, 4, 0, 0, 0, 0], "nao1": [19, 6, 20, 1, 0, 0, 0, 0], "zhai2": [31, 13, 6, 14, 2, 0, 0, 0], "che4": [8, 13, 10, 4, 0, 0, 0, 0], "shou3": [24, 13, 20, 26, 3, 0, 0, 0], "song4": [24, 20, 19, 12, 4, 0, 0, 0], "kuan1": [16, 26, 6, 19, 1, 0, 0, 0], "zong1": [31, 20, 19, 12, 1, 0, 0, 0], "chong3": [8, 13, 20, 19, 12, 3, 0, 0], "xiong4": [29, 14, 20, 19, 12, 4, 0, 0], "jia": [15, 14, 6, 5, 0, 0, 0, 0], "rong2": [23, 20, 19, 12, 2, 0, 0, 0], "xiu3": [29, 14, 26, 3, 0, 0, 0, 0], "cun3": [8, 26, 19, 3, 0, 0, 0, 0], "xun2": [29, 26, 19, 2, 0, 0, 0, 0], "xin2": [29, 14, 19, 2, 0, 0, 0, 0], "she4": [24, 13, 10, 4, 0, 0, 0, 0], "shuan4": [24, 13, 26, 6, 19, 4, 0, 0], "xiao3": [29, 14, 6, 20, 3, 0, 0, 0], "shao3": [24, 13, 6, 20, 3, 0, 0, 0], "er3": [10, 23, 3, 0, 0, 0, 0, 0], "ga4": [12, 6, 4, 0, 0, 0, 0, 0], "che3": [8, 13, 10, 3, 0, 0, 0, 0], "kao1": [16, 6, 20, 1, 0, 0, 0, 0], "niao4": [19, 14, 6, 20, 4, 0, 0, 0], "sui1": [24, 26, 14, 1, 0, 0, 0, 0], "zhan3": [31, 13, 6, 19, 3, 0, 0, 0], "shu3": [24, 13, 26, 3, 0, 0, 0, 0], "zhun1": [31, 13, 26, 19, 1, 0, 0, 0], "sui4": [24, 26, 14, 4, 0, 0, 0, 0], "cen2": [8, 10, 19, 2, 0, 0, 0, 0], "gang3": [12, 6, 19, 12, 3, 0, 0, 0], "min2": [18, 14, 19, 2, 0, 0, 0, 0], "an4": [6, 19, 4, 0, 0, 0, 0, 0], "kong1": [16, 20, 19, 12, 1, 0, 0, 0], "chong2": [8, 13, 20, 19, 12, 2, 0, 0], "wai3": [28, 6, 14, 3, 0, 0, 0, 0], "cuo2": [8, 26, 20, 2, 0, 0, 0, 0], "chuan1": [8, 13, 26, 6, 19, 1, 0, 0], "qiao3": [22, 14, 6, 20, 3, 0, 0, 0], "chai1": [8, 13, 6, 14, 1, 0, 0, 0], "chai4": [8, 13, 6, 14, 4, 0, 0, 0], "cuo1": [8, 26, 20, 1, 0, 0, 0, 0], "shuai4": [24, 13, 26, 6, 14, 4, 0, 0], "fan1": [11, 6, 19, 1, 0, 0, 0, 0], "pa4": [21, 6, 4, 0, 0, 0, 0, 0], "tie4": [25, 14, 10, 4, 0, 0, 0, 0], "tie3": [25, 14, 10, 3, 0, 0, 0, 0], "zhou3": [31, 13, 20, 26, 3, 0, 0, 0], "bang1": [7, 6, 19, 12, 1, 0, 0, 0], "huang3": [13, 26, 6, 19, 12, 3, 0, 0], "chuang2": [8, 13, 26, 6, 19, 12, 2, 0], "nian2": [19, 14, 6, 19, 2, 0, 0, 0], "guan3": [12, 26, 6, 19, 3, 0, 0, 0], "guang3": [12, 26, 6, 19, 12, 3, 0, 0], "ku4": [16, 26, 4, 0, 0, 0, 0, 0], "xiang2": [29, 14, 6, 19, 12, 2, 0, 0], "zhe1": [31, 13, 10, 1, 0, 0, 0, 0], "kang1": [16, 6, 19, 12, 1, 0, 0, 0], "yong2": [30, 20, 19, 12, 2, 0, 0, 0], "lang2": [17, 6, 19, 12, 2, 0, 0, 0], "kuo4": [16, 26, 20, 4, 0, 0, 0, 0], "kai1": [16, 6, 14, 1, 0, 0, 0, 0], "long4": [17, 20, 19, 12, 4, 0, 0, 0], "tui2": [25, 26, 14, 2, 0, 0, 0, 0], "zhang1": [31, 13, 6, 19, 12, 1, 0, 0], "mi2": [18, 14, 2, 0, 0, 0, 0, 0], "qiang3": [22, 14, 6, 19, 12, 3, 0, 0], "cai3": [8, 6, 14, 3, 0, 0, 0, 0], "wang3": [28, 6, 19, 12, 3, 0, 0, 0], "dei3": [9, 10, 14, 3, 0, 0, 0, 0], "zong4": [31, 20, 19, 12, 4, 0, 0, 0], "zong3": [31, 20, 19, 12, 3, 0, 0, 0], "xin1": [29, 14, 19, 1, 0, 0, 0, 0], "ren3": [23, 10, 19, 3, 0, 0, 0, 0], "chan4": [8, 13, 6, 19, 4, 0, 0, 0], "qian3": [22, 14, 6, 19, 3, 0, 0, 0], "keng3": [16, 10, 19, 12, 3, 0, 0, 0], "tui1": [25, 26, 14, 1, 0, 0, 0, 0], "tei1": [25, 10, 14, 1, 0, 0, 0, 0], "mang2": [18, 6, 19, 12, 2, 0, 0, 0], "song3": [24, 20, 19, 12, 3, 0, 0, 0], "zen3": [31, 10, 19, 3, 0, 0, 0, 0], "yang4": [30, 6, 19, 12, 4, 0, 0, 0], "nu4": [19, 26, 4, 0, 0, 0, 0, 0], "guai4": [12, 26, 6, 14, 4, 0, 0, 0], "nin2": [19, 14, 19, 2, 0, 0, 0, 0], "heng2": [13, 10, 19, 12, 2, 0, 0, 0], "lian4": [17, 14, 6, 19, 4, 0, 0, 0], "hen4": [13, 10, 19, 4, 0, 0, 0, 0], "en1": [10, 19, 1, 0, 0, 0, 0, 0], "nao3": [19, 6, 20, 3, 0, 0, 0, 0], "qiao1": [22, 14, 6, 20, 1, 0, 0, 0], "hui3": [13, 26, 14, 3, 0, 0, 0, 0], "min3": [18, 14, 19, 3, 0, 0, 0, 0], "men1": [18, 10, 19, 1, 0, 0, 0, 0], "qing2": [22, 14, 19, 12, 2, 0, 0, 0], "can3": [8, 6, 19, 3, 0, 0, 0, 0], "can2": [8, 6, 19, 2, 0, 0, 0, 0], "gan3": [12, 6, 19, 3, 0, 0, 0, 0], "leng4": [17, 10, 19, 12, 4, 0, 0, 0], "huang1": [13, 26, 6, 19, 12, 1, 0, 0], "huang": [13, 26, 6, 19, 12, 5, 0, 0], "shen4": [24, 13, 10, 19, 4, 0, 0, 0], "bie1": [7, 14, 10, 1, 0, 0, 0, 0], "han1": [13, 6, 19, 1, 0, 0, 0, 0], "dong3": [9, 20, 19, 12, 3, 0, 0, 0], "qu": [22, 26, 5, 0, 0, 0, 0, 0], "reng1": [23, 10, 19, 12, 1, 0, 0, 0], "zang1": [31, 6, 19, 12, 1, 0, 0, 0], "chuo1": [8, 13, 26, 20, 1, 0, 0, 0], "cai2": [8, 6, 14, 2, 0, 0, 0, 0], "zha3": [31, 13, 6, 3, 0, 0, 0, 0], "pa2": [21, 6, 2, 0, 0, 0, 0, 0], "da3": [9, 6, 3, 0, 0, 0, 0, 0], "reng4": [23, 10, 19, 12, 4, 0, 0, 0], "kang2": [16, 6, 19, 12, 2, 0, 0, 0], "niu3": [19, 14, 26, 3, 0, 0, 0, 0], "fen3": [11, 10, 19, 3, 0, 0, 0, 0], "zhao3": [31, 13, 6, 20, 3, 0, 0, 0], "ba3": [7, 6, 3, 0, 0, 0, 0, 0], "zhua1": [31, 13, 26, 6, 1, 0, 0, 0], "dou3": [9, 20, 26, 3, 0, 0, 0, 0], "pao1": [21, 6, 20, 1, 0, 0, 0, 0], "kou1": [16, 20, 26, 1, 0, 0, 0, 0], "lun1": [17, 26, 19, 1, 0, 0, 0, 0], "mo3": [18, 20, 3, 0, 0, 0, 0, 0], "chou1": [8, 13, 20, 26, 1, 0, 0, 0], "nian1": [19, 14, 6, 19, 1, 0, 0, 0], "nian3": [19, 14, 6, 19, 3, 0, 0, 0], "pai1": [21, 6, 14, 1, 0, 0, 0, 0], "lin1": [17, 14, 19, 1, 0, 0, 0, 0], "ling1": [17, 14, 19, 12, 1, 0, 0, 0], "guai3": [12, 26, 6, 14, 3, 0, 0, 0], "ao3": [6, 20, 3, 0, 0, 0, 0, 0], "niu4": [19, 14, 26, 4, 0, 0, 0, 0], "pin1": [21, 14, 19, 1, 0, 0, 0, 0], "ning3": [19, 14, 19, 12, 3, 0, 0, 0], "shuan1": [24, 13, 26, 6, 19, 1, 0, 0], "kao3": [16, 6, 20, 3, 0, 0, 0, 0], "zhuai1": [31, 13, 26, 6, 14, 1, 0, 0], "zhuai4": [31, 13, 26, 6, 14, 4, 0, 0], "na2": [19, 6, 2, 0, 0, 0, 0, 0], "tiao3": [25, 14, 6, 20, 3, 0, 0, 0], "tiao": [25, 14, 6, 20, 5, 0, 0, 0], "nao2": [19, 6, 20, 2, 0, 0, 0, 0], "ting3": [25, 14, 19, 12, 3, 0, 0, 0], "kun3": [16, 26, 19, 3, 0, 0, 0, 0], "hun2": [13, 26, 19, 2, 0, 0, 0, 0], "shao1": [24, 13, 6, 20, 1, 0, 0, 0], "nie1": [19, 14, 10, 1, 0, 0, 0, 0], "lao1": [17, 6, 20, 1, 0, 0, 0, 0], "sun3": [24, 26, 19, 3, 0, 0, 0, 0], "peng3": [21, 10, 19, 12, 3, 0, 0, 0], "she3": [24, 13, 10, 3, 0, 0, 0, 0], "hen2": [13, 10, 19, 2, 0, 0, 0, 0], "zhang3": [31, 13, 6, 19, 12, 3, 0, 0], "qia1": [22, 14, 6, 1, 0, 0, 0, 0], "pai3": [21, 6, 14, 3, 0, 0, 0, 0], "lve4": [17, 27, 10, 4, 0, 0, 0, 0], "lve3": [17, 27, 10, 3, 0, 0, 0, 0], "kong4": [16, 20, 19, 12, 4, 0, 0, 0], "bai1": [7, 6, 14, 1, 0, 0, 0, 0], "shan3": [24, 13, 6, 19, 3, 0, 0, 0], "rou2": [23, 20, 26, 2, 0, 0, 0, 0], "miao2": [18, 14, 6, 20, 2, 0, 0, 0], "chuai1": [8, 13, 26, 6, 14, 1, 0, 0], "chuai3": [8, 13, 26, 6, 14, 3, 0, 0], "zhui1": [31, 13, 26, 14, 1, 0, 0, 0], "ye2": [30, 10, 2, 0, 0, 0, 0, 0], "lou1": [17, 20, 26, 1, 0, 0, 0, 0], "cuo3": [8, 26, 20, 3, 0, 0, 0, 0], "gao3": [12, 6, 20, 3, 0, 0, 0, 0], "kao4": [16, 6, 20, 4, 0, 0, 0, 0], "en4": [10, 19, 4, 0, 0, 0, 0, 0], "bin4": [7, 14, 19, 4, 0, 0, 0, 0], "shuai1": [24, 13, 26, 6, 14, 1, 0, 0], "zui4": [31, 26, 14, 4, 0, 0, 0, 0], "mo1": [18, 20, 1, 0, 0, 0, 0, 0], "sen1": [24, 10, 19, 1, 0, 0, 0, 0], "pie1": [21, 14, 10, 1, 0, 0, 0, 0], "sa3": [24, 6, 3, 0, 0, 0, 0, 0], "rao4": [23, 6, 20, 4, 0, 0, 0, 0], "liao1": [17, 14, 6, 20, 1, 0, 0, 0], "bo3": [7, 20, 3, 0, 0, 0, 0, 0], "zuan1": [31, 26, 6, 19, 1, 0, 0, 0], "chua1": [8, 13, 26, 6, 1, 0, 0, 0], "suan4": [24, 26, 6, 19, 4, 0, 0, 0], "cao1": [8, 6, 20, 1, 0, 0, 0, 0], "bo4": [7, 20, 4, 0, 0, 0, 0, 0], "zan3": [31, 6, 19, 3, 0, 0, 0, 0], "cuan2": [8, 26, 6, 19, 2, 0, 0, 0], "rang4": [23, 6, 19, 12, 4, 0, 0, 0], "luo3": [17, 26, 20, 3, 0, 0, 0, 0], "zan4": [31, 6, 19, 4, 0, 0, 0, 0], "nan4": [19, 6, 19, 4, 0, 0, 0, 0], "zuan4": [31, 26, 6, 19, 4, 0, 0, 0], "gai3": [12, 6, 14, 3, 0, 0, 0, 0], "fang4": [11, 6, 19, 12, 4, 0, 0, 0], "lian3": [17, 14, 6, 19, 3, 0, 0, 0], "cheng4": [8, 13, 10, 19, 12, 4, 0, 0], "san4": [24, 6, 19, 4, 0, 0, 0, 0], "wen2": [28, 10, 19, 2, 0, 0, 0, 0], "zhe4": [31, 13, 10, 4, 0, 0, 0, 0], "duan4": [9, 26, 6, 19, 4, 0, 0, 0], "ri4": [23, 14, 4, 0, 0, 0, 0, 0], "zao3": [31, 6, 20, 3, 0, 0, 0, 0], "hun4": [13, 26, 19, 4, 0, 0, 0, 0], "chun1": [8, 13, 26, 19, 1, 0, 0, 0], "chun3": [8, 13, 26, 19, 3, 0, 0, 0], "huang4": [13, 26, 6, 19, 12, 4, 0, 0], "shai4": [24, 13, 6, 14, 4, 0, 0, 0], "yun1": [30, 26, 19, 1, 0, 0, 0, 0], "gui3": [12, 26, 14, 3, 0, 0, 0, 0], "nuan3": [19, 26, 6, 19, 3, 0, 0, 0], "yue1": [30, 26, 10, 1, 0, 0, 0, 0], "lang3": [17, 6, 19, 12, 3, 0, 0, 0], "mang3": [18, 6, 19, 12, 3, 0, 0, 0], "ben3": [7, 10, 19, 3, 0, 0, 0, 0], "zhu2": [31, 13, 26, 2, 0, 0, 0, 0], "po4": [21, 20, 4, 0, 0, 0, 0, 0], "cun1": [8, 26, 19, 1, 0, 0, 0, 0], "shuo2": [24, 13, 26, 20, 2, 0, 0, 0], "gang4": [12, 6, 19, 12, 4, 0, 0, 0], "gou3": [12, 20, 26, 3, 0, 0, 0, 0], "mou3": [18, 20, 26, 3, 0, 0, 0, 0], "zhou2": [31, 13, 20, 26, 2, 0, 0, 0], "liu3": [17, 14, 26, 3, 0, 0, 0, 0], "gen1": [12, 10, 19, 1, 0, 0, 0, 0], "gun4": [12, 26, 19, 4, 0, 0, 0, 0], "leng2": [17, 10, 19, 12, 2, 0, 0, 0], "leng1": [17, 10, 19, 12, 1, 0, 0, 0], "kuan3": [16, 26, 6, 19, 3, 0, 0, 0], "peng4": [21, 10, 19, 12, 4, 0, 0, 0], "mei3": [18, 10, 14, 3, 0, 0, 0, 0], "bang3": [7, 6, 19, 12, 3, 0, 0, 0], "zao1": [31, 6, 20, 1, 0, 0, 0, 0], "biao4": [7, 14, 6, 20, 4, 0, 0, 0], "mu2": [18, 26, 2, 0, 0, 0, 0, 0], "heng4": [13, 10, 19, 12, 4, 0, 0, 0], "huan1": [13, 26, 6, 19, 1, 0, 0, 0], "ê2": [0, 0, 0, 0, 0, 0, 0, 0], "ei2": [10, 14, 2, 0, 0, 0, 0, 0], "ê3": [0, 0, 0, 0, 0, 0, 0, 0], "ei3": [10, 14, 3, 0, 0, 0, 0, 0], "ê4": [0, 0, 0, 0, 0, 0, 0, 0], "ei4": [10, 14, 4, 0, 0, 0, 0, 0], "ê1": [0, 0, 0, 0, 0, 0, 0, 0], "ei1": [10, 14, 1, 0, 0, 0, 0, 0], "qin4": [22, 14, 19, 4, 0, 0, 0, 0], "ci3": [8, 14, 3, 0, 0, 0, 0, 0], "si3": [24, 14, 3, 0, 0, 0, 0, 0], "du2": [9, 26, 2, 0, 0, 0, 0, 0], "shui3": [24, 13, 26, 14, 3, 0, 0, 0], "ting4": [25, 14, 19, 12, 4, 0, 0, 0], "cuan1": [8, 26, 6, 19, 1, 0, 0, 0], "que1": [22, 26, 10, 1, 0, 0, 0, 0], "xue4": [29, 26, 10, 4, 0, 0, 0, 0], "pen2": [21, 10, 19, 2, 0, 0, 0, 0], "fa3": [11, 6, 3, 0, 0, 0, 0, 0], "feng3": [11, 10, 19, 12, 3, 0, 0, 0], "pao4": [21, 6, 20, 4, 0, 0, 0, 0], "beng4": [7, 10, 19, 12, 4, 0, 0, 0], "sen3": [24, 10, 19, 3, 0, 0, 0, 0], "cui3": [8, 26, 14, 3, 0, 0, 0, 0], "guo1": [12, 26, 20, 1, 0, 0, 0, 0], "lang4": [17, 6, 19, 12, 4, 0, 0, 0], "hai3": [13, 6, 14, 3, 0, 0, 0, 0], "run4": [23, 26, 19, 4, 0, 0, 0, 0], "pou2": [21, 20, 26, 2, 0, 0, 0, 0], "nao4": [19, 6, 20, 4, 0, 0, 0, 0], "shuang4": [24, 13, 26, 6, 19, 12, 4, 0], "lun3": [17, 26, 19, 3, 0, 0, 0, 0], "tian4": [25, 14, 6, 19, 4, 0, 0, 0], "tuan1": [25, 26, 6, 19, 1, 0, 0, 0], "tang4": [25, 6, 19, 12, 4, 0, 0, 0], "man3": [18, 6, 19, 3, 0, 0, 0, 0], "liu1": [17, 14, 26, 1, 0, 0, 0, 0], "ming3": [18, 14, 19, 12, 3, 0, 0, 0], "chou4": [8, 13, 20, 26, 4, 0, 0, 0], "mie4": [18, 14, 10, 4, 0, 0, 0, 0], "teng2": [25, 10, 19, 12, 2, 0, 0, 0], "lan4": [17, 6, 19, 4, 0, 0, 0, 0], "piao3": [21, 14, 6, 20, 3, 0, 0, 0], "lou4": [17, 20, 26, 4, 0, 0, 0, 0], "cao4": [8, 6, 20, 4, 0, 0, 0, 0], "ruan3": [23, 26, 6, 19, 3, 0, 0, 0], "nuan2": [19, 26, 6, 19, 2, 0, 0, 0], "chen4": [8, 13, 10, 19, 4, 0, 0, 0], "ran2": [23, 6, 19, 2, 0, 0, 0, 0], "xiong2": [29, 14, 20, 19, 12, 2, 0, 0], "shou2": [24, 13, 20, 26, 2, 0, 0, 0], "zhua3": [31, 13, 26, 6, 3, 0, 0, 0], "shuang3": [24, 13, 26, 6, 19, 12, 3, 0], "pian4": [21, 14, 6, 19, 4, 0, 0, 0], "niu2": [19, 14, 26, 2, 0, 0, 0, 0], "hang3": [13, 6, 19, 12, 3, 0, 0, 0], "cai1": [8, 6, 14, 1, 0, 0, 0, 0], "mao1": [18, 6, 20, 1, 0, 0, 0, 0], "wa4": [28, 6, 4, 0, 0, 0, 0, 0], "su1": [24, 26, 1, 0, 0, 0, 0, 0], "shuai3": [24, 13, 26, 6, 14, 3, 0, 0], "beng2": [7, 10, 19, 12, 2, 0, 0, 0], "dang": [9, 6, 19, 12, 5, 0, 0, 0], "yuan3": [30, 26, 6, 19, 3, 0, 0, 0], "nve4": [19, 27, 10, 4, 0, 0, 0, 0], "suan1": [24, 26, 6, 19, 1, 0, 0, 0], "tuan3": [25, 26, 6, 19, 3, 0, 0, 0], "fei2": [11, 10, 14, 2, 0, 0, 0, 0], "bie3": [7, 14, 10, 3, 0, 0, 0, 0], "que2": [22, 26, 10, 2, 0, 0, 0, 0], "bai2": [7, 6, 14, 2, 0, 0, 0, 0], "ang4": [6, 19, 12, 4, 0, 0, 0, 0], "dun3": [9, 26, 19, 3, 0, 0, 0, 0], "xing3": [29, 14, 19, 12, 3, 0, 0, 0], "zhe": [31, 13, 10, 5, 0, 0, 0, 0], "zhao2": [31, 13, 6, 20, 2, 0, 0, 0], "shui4": [24, 13, 26, 14, 4, 0, 0, 0], "meng1": [18, 10, 19, 12, 1, 0, 0, 0], "duan3": [9, 26, 6, 19, 3, 0, 0, 0], "che1": [8, 13, 10, 1, 0, 0, 0, 0], "tui4": [25, 26, 14, 4, 0, 0, 0, 0], "tuan4": [25, 26, 6, 19, 4, 0, 0, 0], "qiong1": [22, 14, 20, 19, 12, 1, 0, 0], "zhai3": [31, 13, 6, 14, 3, 0, 0, 0], "cuan4": [8, 26, 6, 19, 4, 0, 0, 0], "neng2": [19, 10, 19, 12, 2, 0, 0, 0], "duan1": [9, 26, 6, 19, 1, 0, 0, 0], "ti": [25, 14, 5, 0, 0, 0, 0, 0], "deng3": [9, 10, 19, 12, 3, 0, 0, 0], "shai1": [24, 13, 6, 14, 1, 0, 0, 0], "zan1": [31, 6, 19, 1, 0, 0, 0, 0], "cu1": [8, 26, 1, 0, 0, 0, 0, 0], "qiu3": [22, 14, 26, 3, 0, 0, 0, 0], "gei3": [12, 10, 14, 3, 0, 0, 0, 0], "sui2": [24, 26, 14, 2, 0, 0, 0, 0], "rui2": [23, 26, 14, 2, 0, 0, 0, 0], "huan3": [13, 26, 6, 19, 3, 0, 0, 0], "sheng2": [24, 13, 10, 19, 12, 2, 0, 0], "miu4": [18, 14, 26, 4, 0, 0, 0, 0], "zuan3": [31, 26, 6, 19, 3, 0, 0, 0], "qun2": [22, 26, 19, 2, 0, 0, 0, 0], "shua3": [24, 13, 26, 6, 3, 0, 0, 0], "nou4": [19, 20, 26, 4, 0, 0, 0, 0], "ping4": [21, 14, 19, 12, 4, 0, 0, 0], "rou4": [23, 20, 26, 4, 0, 0, 0, 0], "ang1": [6, 19, 12, 1, 0, 0, 0, 0], "pang4": [21, 6, 19, 12, 4, 0, 0, 0], "nai2": [19, 6, 14, 2, 0, 0, 0, 0], "tui3": [25, 26, 14, 3, 0, 0, 0, 0], "pang3": [21, 6, 19, 12, 3, 0, 0, 0], "cang2": [8, 6, 19, 12, 2, 0, 0, 0], "gen3": [12, 10, 19, 3, 0, 0, 0, 0], "shai3": [24, 13, 6, 14, 3, 0, 0, 0], "cao3": [8, 6, 20, 3, 0, 0, 0, 0], "zou1": [31, 20, 26, 1, 0, 0, 0, 0], "re2": [23, 10, 2, 0, 0, 0, 0, 0], "ku3": [16, 26, 3, 0, 0, 0, 0, 0], "rong1": [23, 20, 19, 12, 1, 0, 0, 0], "bi2": [7, 14, 2, 0, 0, 0, 0, 0], "cai4": [8, 6, 14, 4, 0, 0, 0, 0], "cang3": [8, 6, 19, 12, 3, 0, 0, 0], "hao1": [13, 6, 20, 1, 0, 0, 0, 0], "xu": [29, 26, 5, 0, 0, 0, 0, 0], "rui3": [23, 26, 14, 3, 0, 0, 0, 0], "ha2": [13, 6, 2, 0, 0, 0, 0, 0], "niao3": [19, 14, 6, 20, 3, 0, 0, 0], "shang": [24, 13, 6, 19, 12, 5, 0, 0], "tun4": [25, 26, 19, 4, 0, 0, 0, 0], "shuo1": [24, 13, 26, 20, 1, 0, 0, 0], "shui2": [24, 13, 26, 14, 2, 0, 0, 0], "shei2": [24, 13, 10, 14, 2, 0, 0, 0], "tou3": [25, 20, 26, 3, 0, 0, 0, 0], "zei2": [31, 10, 14, 2, 0, 0, 0, 0], "zou3": [31, 20, 26, 3, 0, 0, 0, 0], "cou3": [8, 20, 26, 3, 0, 0, 0, 0], "pao3": [21, 6, 20, 3, 0, 0, 0, 0], "zhuai3": [31, 13, 26, 6, 14, 3, 0, 0], "rou3": [23, 20, 26, 3, 0, 0, 0, 0], "ceng4": [8, 10, 19, 12, 4, 0, 0, 0], "zun2": [31, 26, 19, 2, 0, 0, 0, 0], "qun3": [22, 26, 19, 3, 0, 0, 0, 0], "jiu": [15, 14, 26, 5, 0, 0, 0, 0], "jue3": [15, 26, 10, 3, 0, 0, 0, 0], "zhuan3": [31, 13, 26, 6, 19, 3, 0, 0], "bian": [7, 14, 6, 19, 5, 0, 0, 0], "zhei4": [31, 13, 10, 14, 4, 0, 0, 0], "tou4": [25, 20, 26, 4, 0, 0, 0, 0], "qun1": [22, 26, 19, 1, 0, 0, 0, 0], "guo": [12, 26, 20, 5, 0, 0, 0, 0], "niang4": [19, 14, 6, 19, 12, 4, 0, 0], "pan3": [21, 6, 19, 3, 0, 0, 0, 0], "chuang3": [8, 13, 26, 6, 19, 12, 3, 0], "long1": [17, 20, 19, 12, 1, 0, 0, 0], "xue3": [29, 26, 10, 3, 0, 0, 0, 0], "zhun4": [31, 13, 26, 19, 4, 0, 0, 0], "gu2": [12, 26, 2, 0, 0, 0, 0, 0], "zang3": [31, 6, 19, 12, 3, 0, 0, 0], "sui3": [24, 26, 14, 3, 0, 0, 0, 0], "song2": [24, 20, 19, 12, 2, 0, 0, 0], "hou1": [13, 20, 26, 1, 0, 0, 0, 0], "tai3": [25, 6, 14, 3, 0, 0, 0, 0], "jiang": [15, 14, 6, 19, 12, 5, 0, 0], "shui": [24, 13, 26, 14, 5, 0, 0, 0], "pou4": [21, 20, 26, 4, 0, 0, 0, 0], "zun3": [31, 26, 19, 3, 0, 0, 0, 0], "mou1": [18, 20, 26, 1, 0, 0, 0, 0], "pian3": [21, 14, 6, 19, 3, 0, 0, 0], "nang3": [19, 6, 19, 12, 3, 0, 0, 0], "xin3": [29, 14, 19, 3, 0, 0, 0, 0], "xiong3": [29, 14, 20, 19, 12, 3, 0, 0], "mang1": [18, 6, 19, 12, 1, 0, 0, 0], "chai3": [8, 13, 6, 14, 3, 0, 0, 0], "den4": [9, 10, 19, 4, 0, 0, 0, 0], "chen": [8, 13, 10, 19, 5, 0, 0, 0], "an2": [6, 19, 2, 0, 0, 0, 0, 0], "pei3": [21, 10, 14, 3, 0, 0, 0, 0], "kuai3": [16, 26, 6, 14, 3, 0, 0, 0], "sun4": [24, 26, 19, 4, 0, 0, 0, 0], "nuo3": [19, 26, 20, 3, 0, 0, 0, 0], "hun3": [13, 26, 19, 3, 0, 0, 0, 0], "kuang3": [16, 26, 6, 19, 12, 3, 0, 0], "nie2": [19, 14, 10, 2, 0, 0, 0, 0], "fou2": [11, 20, 26, 2, 0, 0, 0, 0], "qiu4": [22, 14, 26, 4, 0, 0, 0, 0]}

config/pinyin_map.json

@@ -0,0 +1 @@
+{"idx2char": ["0", "1", "2", "3", "4", "5", "a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z"], "char2idx": {"0": 0, "1": 1, "2": 2, "3": 3, "4": 4, "5": 5, "a": 6, "b": 7, "c": 8, "d": 9, "e": 10, "f": 11, "g": 12, "h": 13, "i": 14, "j": 15, "k": 16, "l": 17, "m": 18, "n": 19, "o": 20, "p": 21, "q": 22, "r": 23, "s": 24, "t": 25, "u": 26, "v": 27, "w": 28, "x": 29, "y": 30, "z": 31}}

config/方正古隶繁体.ttf24.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b354344e73ed5afc3d9a922d9cef2fc6f53bfff17a5876b3ffb74e74df02b4f
+size 107071616

csc_model.py

@@ -0,0 +1,519 @@
+import json
+import os
+import shutil
+import time
+from pathlib import Path
+from typing import List
+
+import numpy as np
+import torch
+from huggingface_hub import hf_hub_download
+from huggingface_hub.file_download import http_user_agent
+from torch import nn
+from torch.nn import functional as F
+from transformers import BertPreTrainedModel, BertModel
+from transformers.modeling_outputs import MaskedLMOutput, BaseModelOutputWithPooling
+from transformers.models.bert.modeling_bert import BertEncoder, BertPooler, BertLMPredictionHead
+
+cache_path = Path(os.path.abspath(__file__)).parent
+
+
+def download_file(filename: str, path: Path):
+    if os.path.exists(cache_path / filename):
+        return
+
+    if os.path.exists(path / filename):
+        shutil.copyfile(path / filename, cache_path / filename)
+        return
+
+    hf_hub_download(
+        "iioSnail/ChineseBERT-for-csc",
+        filename,
+        local_dir=cache_path,
+        user_agent=http_user_agent(),
+    )
+    time.sleep(0.2)
+
+
+class ChineseBertForCSC(BertPreTrainedModel):
+
+    def __init__(self, config):
+        super(ChineseBertForCSC, self).__init__(config)
+        self.model = Dynamic_GlyceBertForMultiTask(config)
+        self.tokenizer = None
+
+    def forward(self, **kwargs):
+        return self.model(**kwargs)
+
+    def set_tokenizer(self, tokenizer):
+        self.tokenizer = tokenizer
+
+    def _predict(self, sentence):
+        if self.tokenizer is None:
+            return "Please init tokenizer by `set_tokenizer(tokenizer)` before predict."
+
+        inputs = self.tokenizer([sentence], return_tensors='pt')
+        output_hidden = self.model(**inputs).logits
+        return self.tokenizer.convert_ids_to_tokens(output_hidden.argmax(-1)[0, 1:-1])
+
+    def predict(self, sentence, window=1):
+        _src_tokens = list(sentence)
+        src_tokens = list(sentence)
+        pred_tokens = self._predict(sentence)
+
+        for _ in range(window):
+            record_index = []
+            for i, (a, b) in enumerate(zip(src_tokens, pred_tokens)):
+                if a != b:
+                    record_index.append(i)
+
+            src_tokens = pred_tokens
+            pred_tokens = self._predict(''.join(pred_tokens))
+            for i, (a, b) in enumerate(zip(src_tokens, pred_tokens)):
+                # 若这个token被修改了，且在窗口范围内，则什么都不做。
+                if a != b and any([abs(i - x) <= 1 for x in record_index]):
+                    pass
+                else:
+                    pred_tokens[i] = src_tokens[i]
+
+        return ''.join(pred_tokens)
+
+
+#################################ChineseBERT Source Code##############################################
+class Dynamic_GlyceBertForMultiTask(BertPreTrainedModel):
+    def __init__(self, config):
+        super(Dynamic_GlyceBertForMultiTask, self).__init__(config)
+
+        self.bert = GlyceBertModel(config)
+        self.cls = MultiTaskHeads(config)
+
+    def get_output_embeddings(self):
+        return self.cls.predictions.decoder
+
+    def forward(
+            self,
+            input_ids=None,
+            pinyin_ids=None,
+            attention_mask=None,
+            token_type_ids=None,
+            position_ids=None,
+            head_mask=None,
+            inputs_embeds=None,
+            encoder_hidden_states=None,
+            encoder_attention_mask=None,
+            output_attentions=None,
+            output_hidden_states=None,
+            return_dict=None,
+            **kwargs
+    ):
+        assert kwargs == {}, f"Unexpected keyword arguments: {list(kwargs.keys())}."
+
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        outputs_x = self.bert(
+            input_ids,
+            pinyin_ids,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        encoded_x = outputs_x[0]
+
+        prediction_scores = self.cls(encoded_x)
+
+        return MaskedLMOutput(
+            logits=prediction_scores,
+            hidden_states=outputs_x.hidden_states,
+            attentions=outputs_x.attentions,
+        )
+
+
+class GlyceBertModel(BertModel):
+    r"""
+    Outputs: `Tuple` comprising various elements depending on the configuration (config) and inputs:
+        **last_hidden_state**: ``torch.FloatTensor`` of shape ``(batch_size, sequence_length, hidden_size)``
+            Sequence of hidden-states at the output of the last layer of the models.
+        **pooler_output**: ``torch.FloatTensor`` of shape ``(batch_size, hidden_size)``
+            Last layer hidden-state of the first token of the sequence (classification token)
+            further processed by a Linear layer and a Tanh activation function. The Linear
+            layer weights are trained from the next sentence prediction (classification)
+            objective during Bert pretraining. This output is usually *not* a good summary
+            of the semantic content of the input, you're often better with averaging or pooling
+            the sequence of hidden-states for the whole input sequence.
+        **hidden_states**: (`optional`, returned when ``config.output_hidden_states=True``)
+            list of ``torch.FloatTensor`` (one for the output of each layer + the output of the embeddings)
+            of shape ``(batch_size, sequence_length, hidden_size)``:
+            Hidden-states of the models at the output of each layer plus the initial embedding outputs.
+        **attentions**: (`optional`, returned when ``config.output_attentions=True``)
+            list of ``torch.FloatTensor`` (one for each layer) of shape ``(batch_size, num_heads, sequence_length, sequence_length)``:
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.
+
+    Examples::
+
+        tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
+        models = BertModel.from_pretrained('bert-base-uncased')
+        input_ids = torch.tensor(tokenizer.encode("Hello, my dog is cute", add_special_tokens=True)).unsqueeze(0)  # Batch size 1
+        outputs = models(input_ids)
+        last_hidden_states = outputs[0]  # The last hidden-state is the first element of the output tuple
+
+    """
+
+    def __init__(self, config):
+        super(GlyceBertModel, self).__init__(config)
+        self.config = config
+
+        self.embeddings = FusionBertEmbeddings(config)
+        self.encoder = BertEncoder(config)
+        self.pooler = BertPooler(config)
+
+        self.init_weights()
+
+    def forward(
+            self,
+            input_ids=None,
+            pinyin_ids=None,
+            attention_mask=None,
+            token_type_ids=None,
+            position_ids=None,
+            head_mask=None,
+            inputs_embeds=None,
+            encoder_hidden_states=None,
+            encoder_attention_mask=None,
+            output_attentions=None,
+            output_hidden_states=None,
+            return_dict=None,
+    ):
+        r"""
+        encoder_hidden_states  (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`):
+            Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention
+            if the models is configured as a decoder.
+        encoder_attention_mask (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length)`, `optional`):
+            Mask to avoid performing attention on the padding token indices of the encoder input. This mask
+            is used in the cross-attention if the models is configured as a decoder.
+            Mask values selected in ``[0, 1]``:
+
+            - 1 for tokens that are **not masked**,
+            - 0 for tokens that are **masked**.
+        """
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
+        elif input_ids is not None:
+            input_shape = input_ids.size()
+        elif inputs_embeds is not None:
+            input_shape = inputs_embeds.size()[:-1]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        device = input_ids.device if input_ids is not None else inputs_embeds.device
+
+        if attention_mask is None:
+            attention_mask = torch.ones(input_shape, device=device)
+        if token_type_ids is None:
+            token_type_ids = torch.zeros(input_shape, dtype=torch.long, device=device)
+
+        # We can provide a self-attention mask of dimensions [batch_size, from_seq_length, to_seq_length]
+        # ourselves in which case we just need to make it broadcastable to all heads.
+        extended_attention_mask: torch.Tensor = self.get_extended_attention_mask(attention_mask, input_shape, device)
+
+        # If a 2D or 3D attention mask is provided for the cross-attention
+        # we need to make broadcastable to [batch_size, num_heads, seq_length, seq_length]
+        if self.config.is_decoder and encoder_hidden_states is not None:
+            encoder_batch_size, encoder_sequence_length, _ = encoder_hidden_states.size()
+            encoder_hidden_shape = (encoder_batch_size, encoder_sequence_length)
+            if encoder_attention_mask is None:
+                encoder_attention_mask = torch.ones(encoder_hidden_shape, device=device)
+            encoder_extended_attention_mask = self.invert_attention_mask(encoder_attention_mask)
+        else:
+            encoder_extended_attention_mask = None
+
+        # Prepare head mask if needed
+        # 1.0 in head_mask indicate we keep the head
+        # attention_probs has shape bsz x n_heads x N x N
+        # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads]
+        # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length]
+        head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)
+
+        embedding_output = self.embeddings(
+            input_ids=input_ids, pinyin_ids=pinyin_ids, position_ids=position_ids, token_type_ids=token_type_ids,
+            inputs_embeds=inputs_embeds
+        )
+        encoder_outputs = self.encoder(
+            embedding_output,
+            attention_mask=extended_attention_mask,
+            head_mask=head_mask,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_extended_attention_mask,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        sequence_output = encoder_outputs[0]
+        pooled_output = self.pooler(sequence_output) if self.pooler is not None else None
+
+        if not return_dict:
+            return (sequence_output, pooled_output) + encoder_outputs[1:]
+
+        return BaseModelOutputWithPooling(
+            last_hidden_state=sequence_output,
+            pooler_output=pooled_output,
+            hidden_states=encoder_outputs.hidden_states,
+            attentions=encoder_outputs.attentions,
+        )
+
+    def forward_with_embedding(
+            self,
+            input_ids=None,
+            pinyin_ids=None,
+            attention_mask=None,
+            token_type_ids=None,
+            position_ids=None,
+            head_mask=None,
+            inputs_embeds=None,
+            encoder_hidden_states=None,
+            encoder_attention_mask=None,
+            output_attentions=None,
+            output_hidden_states=None,
+            return_dict=None,
+            embedding=None
+    ):
+        r"""
+        encoder_hidden_states  (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`):
+            Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention
+            if the models is configured as a decoder.
+        encoder_attention_mask (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length)`, `optional`):
+            Mask to avoid performing attention on the padding token indices of the encoder input. This mask
+            is used in the cross-attention if the models is configured as a decoder.
+            Mask values selected in ``[0, 1]``:
+
+            - 1 for tokens that are **not masked**,
+            - 0 for tokens that are **masked**.
+        """
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
+        elif input_ids is not None:
+            input_shape = input_ids.size()
+        elif inputs_embeds is not None:
+            input_shape = inputs_embeds.size()[:-1]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        device = input_ids.device if input_ids is not None else inputs_embeds.device
+
+        if attention_mask is None:
+            attention_mask = torch.ones(input_shape, device=device)
+        if token_type_ids is None:
+            token_type_ids = torch.zeros(input_shape, dtype=torch.long, device=device)
+
+        # We can provide a self-attention mask of dimensions [batch_size, from_seq_length, to_seq_length]
+        # ourselves in which case we just need to make it broadcastable to all heads.
+        extended_attention_mask: torch.Tensor = self.get_extended_attention_mask(attention_mask, input_shape, device)
+
+        # If a 2D or 3D attention mask is provided for the cross-attention
+        # we need to make broadcastable to [batch_size, num_heads, seq_length, seq_length]
+        if self.config.is_decoder and encoder_hidden_states is not None:
+            encoder_batch_size, encoder_sequence_length, _ = encoder_hidden_states.size()
+            encoder_hidden_shape = (encoder_batch_size, encoder_sequence_length)
+            if encoder_attention_mask is None:
+                encoder_attention_mask = torch.ones(encoder_hidden_shape, device=device)
+            encoder_extended_attention_mask = self.invert_attention_mask(encoder_attention_mask)
+        else:
+            encoder_extended_attention_mask = None
+
+        # Prepare head mask if needed
+        # 1.0 in head_mask indicate we keep the head
+        # attention_probs has shape bsz x n_heads x N x N
+        # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads]
+        # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length]
+        head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)
+
+        assert embedding is not None
+        embedding_output = embedding
+        encoder_outputs = self.encoder(
+            embedding_output,
+            attention_mask=extended_attention_mask,
+            head_mask=head_mask,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_extended_attention_mask,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        sequence_output = encoder_outputs[0]
+        pooled_output = self.pooler(sequence_output) if self.pooler is not None else None
+
+        if not return_dict:
+            return (sequence_output, pooled_output) + encoder_outputs[1:]
+
+        return BaseModelOutputWithPooling(
+            last_hidden_state=sequence_output,
+            pooler_output=pooled_output,
+            hidden_states=encoder_outputs.hidden_states,
+            attentions=encoder_outputs.attentions,
+        )
+
+
+class MultiTaskHeads(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.predictions = BertLMPredictionHead(config)
+
+    def forward(self, sequence_output):
+        prediction_scores = self.predictions(sequence_output)
+        return prediction_scores
+
+
+class FusionBertEmbeddings(nn.Module):
+    """
+    Construct the embeddings from word, position, glyph, pinyin and token_type embeddings.
+    """
+
+    def __init__(self, config):
+        super(FusionBertEmbeddings, self).__init__()
+
+        self.path = Path(config._name_or_path)
+        config_path = cache_path / 'config'
+        if not os.path.exists(config_path):
+            os.makedirs(config_path)
+
+        font_files = []
+        download_file("config/STFANGSO.TTF24.npy", self.path)
+        download_file("config/STXINGKA.TTF24.npy", self.path)
+        download_file("config/方正古隶繁体.ttf24.npy", self.path)
+        for file in os.listdir(config_path):
+            if file.endswith(".npy"):
+                font_files.append(config_path / file)
+        self.word_embeddings = nn.Embedding(config.vocab_size, config.hidden_size, padding_idx=0)
+        self.position_embeddings = nn.Embedding(config.max_position_embeddings, config.hidden_size)
+        self.token_type_embeddings = nn.Embedding(config.type_vocab_size, config.hidden_size)
+        self.pinyin_embeddings = PinyinEmbedding(embedding_size=128, pinyin_out_dim=config.hidden_size, config=config)
+        self.glyph_embeddings = GlyphEmbedding(font_npy_files=font_files)
+
+        # self.LayerNorm is not snake-cased to stick with TensorFlow models variable name and be able to load
+        # any TensorFlow checkpoint file
+        self.glyph_map = nn.Linear(1728, config.hidden_size)
+        self.map_fc = nn.Linear(config.hidden_size * 3, config.hidden_size)
+        self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.dropout = nn.Dropout(config.hidden_dropout_prob)
+
+        # position_ids (1, len position emb) is contiguous in memory and exported when serialized
+        self.register_buffer("position_ids", torch.arange(config.max_position_embeddings).expand((1, -1)))
+
+    def forward(self, input_ids=None, pinyin_ids=None, token_type_ids=None, position_ids=None, inputs_embeds=None):
+        if input_ids is not None:
+            input_shape = input_ids.size()
+        else:
+            input_shape = inputs_embeds.size()[:-1]
+
+        seq_length = input_shape[1]
+
+        if position_ids is None:
+            position_ids = self.position_ids[:, :seq_length]
+
+        if token_type_ids is None:
+            token_type_ids = torch.zeros(input_shape, dtype=torch.long, device=self.position_ids.device)
+
+        if inputs_embeds is None:
+            inputs_embeds = self.word_embeddings(input_ids)
+
+        # get char embedding, pinyin embedding and glyph embedding
+        word_embeddings = inputs_embeds  # [bs,l,hidden_size]
+        pinyin_embeddings = self.pinyin_embeddings(pinyin_ids)  # [bs,l,hidden_size]
+        glyph_embeddings = self.glyph_map(self.glyph_embeddings(input_ids))  # [bs,l,hidden_size]
+        # fusion layer
+        concat_embeddings = torch.cat((word_embeddings, pinyin_embeddings, glyph_embeddings), 2)
+        inputs_embeds = self.map_fc(concat_embeddings)
+
+        position_embeddings = self.position_embeddings(position_ids)
+        token_type_embeddings = self.token_type_embeddings(token_type_ids)
+
+        embeddings = inputs_embeds + position_embeddings + token_type_embeddings
+        embeddings = self.LayerNorm(embeddings)
+        embeddings = self.dropout(embeddings)
+        return embeddings
+
+
+class PinyinEmbedding(nn.Module):
+
+    def __init__(self, embedding_size: int, pinyin_out_dim: int, config):
+        """
+            Pinyin Embedding Module
+        Args:
+            embedding_size: the size of each embedding vector
+            pinyin_out_dim: kernel number of conv
+        """
+        super(PinyinEmbedding, self).__init__()
+        download_file("config/pinyin_map.json", Path(config._name_or_path))
+        with open(cache_path / 'config' / 'pinyin_map.json') as fin:
+            pinyin_dict = json.load(fin)
+        self.pinyin_out_dim = pinyin_out_dim
+        self.embedding = nn.Embedding(len(pinyin_dict['idx2char']), embedding_size)
+        self.conv = nn.Conv1d(in_channels=embedding_size, out_channels=self.pinyin_out_dim, kernel_size=2,
+                              stride=1, padding=0)
+
+    def forward(self, pinyin_ids):
+        """
+        Args:
+            pinyin_ids: (bs*sentence_length*pinyin_locs)
+
+        Returns:
+            pinyin_embed: (bs,sentence_length,pinyin_out_dim)
+        """
+        # input pinyin ids for 1-D conv
+        embed = self.embedding(pinyin_ids)  # [bs,sentence_length,pinyin_locs,embed_size]
+        bs, sentence_length, pinyin_locs, embed_size = embed.shape
+        view_embed = embed.view(-1, pinyin_locs, embed_size)  # [(bs*sentence_length),pinyin_locs,embed_size]
+        input_embed = view_embed.permute(0, 2, 1)  # [(bs*sentence_length), embed_size, pinyin_locs]
+        # conv + max_pooling
+        pinyin_conv = self.conv(input_embed)  # [(bs*sentence_length),pinyin_out_dim,H]
+        pinyin_embed = F.max_pool1d(pinyin_conv, pinyin_conv.shape[-1])  # [(bs*sentence_length),pinyin_out_dim,1]
+        return pinyin_embed.view(bs, sentence_length, self.pinyin_out_dim)  # [bs,sentence_length,pinyin_out_dim]
+
+
+class GlyphEmbedding(nn.Module):
+    """Glyph2Image Embedding"""
+
+    def __init__(self, font_npy_files: List[str]):
+        super(GlyphEmbedding, self).__init__()
+        font_arrays = [
+            np.load(np_file).astype(np.float32) for np_file in font_npy_files
+        ]
+        self.vocab_size = font_arrays[0].shape[0]
+        self.font_num = len(font_arrays)
+        self.font_size = font_arrays[0].shape[-1]
+        # N, C, H, W
+        font_array = np.stack(font_arrays, axis=1)
+        self.embedding = nn.Embedding(
+            num_embeddings=self.vocab_size,
+            embedding_dim=self.font_size ** 2 * self.font_num,
+            _weight=torch.from_numpy(font_array.reshape([self.vocab_size, -1]))
+        )
+
+    def forward(self, input_ids):
+        """
+            get glyph images for batch inputs
+        Args:
+            input_ids: [batch, sentence_length]
+        Returns:
+            images: [batch, sentence_length, self.font_num*self.font_size*self.font_size]
+        """
+        # return self.embedding(input_ids).view([-1, self.font_num, self.font_size, self.font_size])
+        return self.embedding(input_ids)

csc_tokenizer.py

@@ -0,0 +1,206 @@
+import json
+import os
+import shutil
+import time
+from pathlib import Path
+from typing import List, Union, Optional
+
+import tokenizers
+import torch
+from torch import NoneType
+from huggingface_hub import hf_hub_download
+from huggingface_hub.file_download import http_user_agent
+from pypinyin import pinyin, Style
+from transformers.tokenization_utils_base import TruncationStrategy
+from transformers.utils import PaddingStrategy
+from transformers.utils.generic import TensorType
+
+try:
+    from tokenizers import BertWordPieceTokenizer
+except:
+    from tokenizers.implementations import BertWordPieceTokenizer
+
+from transformers import BertTokenizerFast, BatchEncoding
+
+cache_path = Path(os.path.abspath(__file__)).parent
+
+
+def download_file(filename: str, path: Path):
+    if os.path.exists(cache_path / filename):
+        return
+
+    if os.path.exists(path / filename):
+        shutil.copyfile(path / filename, cache_path / filename)
+        return
+
+    hf_hub_download(
+        "iioSnail/ChineseBERT-for-csc",
+        filename,
+        local_dir=cache_path,
+        user_agent=http_user_agent(),
+    )
+    time.sleep(0.2)
+
+
+class ChineseBertTokenizer(BertTokenizerFast):
+
+    def __init__(self, **kwargs):
+        super(ChineseBertTokenizer, self).__init__(**kwargs)
+
+        self.path = Path(kwargs['name_or_path'])
+        vocab_file = cache_path / 'vocab.txt'
+        config_path = cache_path / 'config'
+        if not os.path.exists(config_path):
+            os.makedirs(config_path)
+
+        self.max_length = 512
+
+        download_file('vocab.txt', self.path)
+        self.tokenizer = BertWordPieceTokenizer(str(vocab_file))
+
+        # load pinyin map dict
+        download_file('config/pinyin_map.json', self.path)
+        with open(config_path / 'pinyin_map.json', encoding='utf8') as fin:
+            self.pinyin_dict = json.load(fin)
+
+        # load char id map tensor
+        download_file('config/id2pinyin.json', self.path)
+        with open(config_path / 'id2pinyin.json', encoding='utf8') as fin:
+            self.id2pinyin = json.load(fin)
+
+        # load pinyin map tensor
+        download_file('config/pinyin2tensor.json', self.path)
+        with open(config_path / 'pinyin2tensor.json', encoding='utf8') as fin:
+            self.pinyin2tensor = json.load(fin)
+
+    def __call__(self,
+                 text: Union[str, List[str], List[List[str]]] = None,
+                 text_pair: Union[str, List[str], List[List[str]], NoneType] = None,
+                 text_target: Union[str, List[str], List[List[str]]] = None,
+                 text_pair_target: Union[str, List[str], List[List[str]], NoneType] = None,
+                 add_special_tokens: bool = True,
+                 padding: Union[bool, str, PaddingStrategy] = False,
+                 truncation: Union[bool, str, TruncationStrategy] = None,
+                 max_length: Optional[int] = None,
+                 stride: int = 0,
+                 is_split_into_words: bool = False,
+                 pad_to_multiple_of: Optional[int] = None,
+                 return_tensors: Union[str, TensorType, NoneType] = None,
+                 return_token_type_ids: Optional[bool] = None,
+                 return_attention_mask: Optional[bool] = None,
+                 return_overflowing_tokens: bool = False, return_special_tokens_mask: bool = False,
+                 return_offsets_mapping: bool = False,
+                 return_length: bool = False,
+                 verbose: bool = True, **kwargs) -> BatchEncoding:
+        encoding = super(ChineseBertTokenizer, self).__call__(
+            text=text,
+            text_pair=text_pair,
+            text_target=text_target,
+            text_pair_target=text_pair_target,
+            add_special_tokens=add_special_tokens,
+            padding=padding,
+            truncation=truncation,
+            max_length=max_length,
+            stride=stride,
+            is_split_into_words=is_split_into_words,
+            pad_to_multiple_of=pad_to_multiple_of,
+            return_tensors=return_tensors,
+            return_token_type_ids=return_token_type_ids,
+            return_attention_mask=return_attention_mask,
+            return_overflowing_tokens=return_overflowing_tokens,
+            return_offsets_mapping=return_offsets_mapping,
+            return_length=return_length,
+            verbose=verbose,
+        )
+
+        input_ids = encoding.input_ids
+
+        pinyin_ids = None
+        if type(text) == str:
+            pinyin_ids = self.convert_ids_to_pinyin_ids(input_ids)
+
+        if type(text) == list:
+            pinyin_ids = []
+            for ids in input_ids:
+                pinyin_ids.append(self.convert_ids_to_pinyin_ids(ids))
+
+        if torch.is_tensor(encoding.input_ids):
+            pinyin_ids = torch.LongTensor(pinyin_ids)
+
+        encoding['pinyin_ids'] = pinyin_ids
+
+        return encoding
+
+    def tokenize_sentence(self, sentence):
+        # convert sentence to ids
+        tokenizer_output = self.tokenizer.encode(sentence)
+        bert_tokens = tokenizer_output.ids
+        pinyin_tokens = self.convert_sentence_to_pinyin_ids(sentence, tokenizer_output)
+        # assert，token nums should be same as pinyin token nums
+        assert len(bert_tokens) <= self.max_length
+        assert len(bert_tokens) == len(pinyin_tokens)
+        # convert list to tensor
+        input_ids = torch.LongTensor(bert_tokens)
+        pinyin_ids = torch.LongTensor(pinyin_tokens).view(-1)
+        return input_ids, pinyin_ids
+
+    def convert_ids_to_pinyin_ids(self, ids: List[int]):
+        pinyin_ids = []
+        tokens = self.convert_ids_to_tokens(ids)
+        for token in tokens:
+            if len(token) > 1:
+                pinyin_ids.append([0] * 8)
+                continue
+
+            pinyin_string = pinyin(token, style=Style.TONE3, errors=lambda x: [['not chinese'] for _ in x])[0][0]
+
+            if pinyin_string == "not chinese":
+                pinyin_ids.append([0] * 8)
+                continue
+
+            if pinyin_string in self.pinyin2tensor:
+                pinyin_ids.append(self.pinyin2tensor[pinyin_string])
+            else:
+                ids = [0] * 8
+                for i, p in enumerate(pinyin_string):
+                    if p not in self.pinyin_dict["char2idx"]:
+                        ids = [0] * 8
+                        break
+                    ids[i] = self.pinyin_dict["char2idx"][p]
+                pinyin_ids.append(pinyin_ids)
+
+        return pinyin_ids
+
+    def convert_sentence_to_pinyin_ids(self, sentence: str, tokenizer_output: tokenizers.Encoding) -> List[List[int]]:
+        # get pinyin of a sentence
+        pinyin_list = pinyin(sentence, style=Style.TONE3, heteronym=True, errors=lambda x: [['not chinese'] for _ in x])
+        pinyin_locs = {}
+        # get pinyin of each location
+        for index, item in enumerate(pinyin_list):
+            pinyin_string = item[0]
+            # not a Chinese character, pass
+            if pinyin_string == "not chinese":
+                continue
+            if pinyin_string in self.pinyin2tensor:
+                pinyin_locs[index] = self.pinyin2tensor[pinyin_string]
+            else:
+                ids = [0] * 8
+                for i, p in enumerate(pinyin_string):
+                    if p not in self.pinyin_dict["char2idx"]:
+                        ids = [0] * 8
+                        break
+                    ids[i] = self.pinyin_dict["char2idx"][p]
+                pinyin_locs[index] = ids
+
+        # find chinese character location, and generate pinyin ids
+        pinyin_ids = []
+        for idx, (token, offset) in enumerate(zip(tokenizer_output.tokens, tokenizer_output.offsets)):
+            if offset[1] - offset[0] != 1:
+                pinyin_ids.append([0] * 8)
+                continue
+            if offset[0] in pinyin_locs:
+                pinyin_ids.append(pinyin_locs[offset[0]])
+            else:
+                pinyin_ids.append([0] * 8)
+
+        return pinyin_ids

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30e02f16f0558b31b692d876637bcd0641db252f9b2d9a9d3f0d6045c6d49cb2
+size 663267797

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,21 @@
+{
+  "auto_map": {
+    "AutoTokenizer": [
+      "csc_tokenizer.ChineseBertTokenizer",
+      null
+    ]
+  },
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": null,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "ChineseBertTokenizer",
+  "unk_token": "[UNK]"
+}