Upload model files.

README.md

@@ -0,0 +1,47 @@
+---
+language:
+- zh
+thumbnail: https://ckip.iis.sinica.edu.tw/files/ckip_logo.png
+tags:
+- pytorch
+- token-classification
+- bert
+- zh
+license: gpl-3.0
+datasets:
+metrics:
+---
+
+# CKIP BERT Tiny Chinese
+
+This project provides traditional Chinese transformers models (including ALBERT, BERT, GPT2) and NLP tools (including word segmentation, part-of-speech tagging, named entity recognition).
+
+這個專案提供了繁體中文的 transformers 模型（包含 ALBERT、BERT、GPT2）及自然語言處理工具（包含斷詞、詞性標記、實體辨識）。
+
+## Homepage
+
+* https://github.com/ckiplab/ckip-transformers
+
+## Contributers
+
+* [Mu Yang](https://muyang.pro) at [CKIP](https://ckip.iis.sinica.edu.tw) (Author & Maintainer)
+
+## Usage
+
+Please use BertTokenizerFast as tokenizer instead of AutoTokenizer.
+
+請使用 BertTokenizerFast 而非 AutoTokenizer。
+
+```
+from transformers import (
+  BertTokenizerFast,
+  AutoModel,
+)
+
+tokenizer = BertTokenizerFast.from_pretrained('bert-base-chinese')
+model = AutoModel.from_pretrained('ckiplab/bert-tiny-chinese-ner')
+```
+
+For full usage and more information, please refer to https://github.com/ckiplab/ckip-transformers.
+
+有關完整使用方法及其他資訊，請參見 https://github.com/ckiplab/ckip-transformers 。

config.json

@@ -0,0 +1,178 @@
+{
+  "_name_or_path": "../../../model/bert-tiny-scratch-lm",
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "directionality": "bidi",
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 312,
+  "id2label": {
+    "0": "O",
+    "1": "B-CARDINAL",
+    "2": "B-DATE",
+    "3": "B-EVENT",
+    "4": "B-FAC",
+    "5": "B-GPE",
+    "6": "B-LANGUAGE",
+    "7": "B-LAW",
+    "8": "B-LOC",
+    "9": "B-MONEY",
+    "10": "B-NORP",
+    "11": "B-ORDINAL",
+    "12": "B-ORG",
+    "13": "B-PERCENT",
+    "14": "B-PERSON",
+    "15": "B-PRODUCT",
+    "16": "B-QUANTITY",
+    "17": "B-TIME",
+    "18": "B-WORK_OF_ART",
+    "19": "I-CARDINAL",
+    "20": "I-DATE",
+    "21": "I-EVENT",
+    "22": "I-FAC",
+    "23": "I-GPE",
+    "24": "I-LANGUAGE",
+    "25": "I-LAW",
+    "26": "I-LOC",
+    "27": "I-MONEY",
+    "28": "I-NORP",
+    "29": "I-ORDINAL",
+    "30": "I-ORG",
+    "31": "I-PERCENT",
+    "32": "I-PERSON",
+    "33": "I-PRODUCT",
+    "34": "I-QUANTITY",
+    "35": "I-TIME",
+    "36": "I-WORK_OF_ART",
+    "37": "E-CARDINAL",
+    "38": "E-DATE",
+    "39": "E-EVENT",
+    "40": "E-FAC",
+    "41": "E-GPE",
+    "42": "E-LANGUAGE",
+    "43": "E-LAW",
+    "44": "E-LOC",
+    "45": "E-MONEY",
+    "46": "E-NORP",
+    "47": "E-ORDINAL",
+    "48": "E-ORG",
+    "49": "E-PERCENT",
+    "50": "E-PERSON",
+    "51": "E-PRODUCT",
+    "52": "E-QUANTITY",
+    "53": "E-TIME",
+    "54": "E-WORK_OF_ART",
+    "55": "S-CARDINAL",
+    "56": "S-DATE",
+    "57": "S-EVENT",
+    "58": "S-FAC",
+    "59": "S-GPE",
+    "60": "S-LANGUAGE",
+    "61": "S-LAW",
+    "62": "S-LOC",
+    "63": "S-MONEY",
+    "64": "S-NORP",
+    "65": "S-ORDINAL",
+    "66": "S-ORG",
+    "67": "S-PERCENT",
+    "68": "S-PERSON",
+    "69": "S-PRODUCT",
+    "70": "S-QUANTITY",
+    "71": "S-TIME",
+    "72": "S-WORK_OF_ART"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 1248,
+  "label2id": {
+    "B-CARDINAL": 1,
+    "B-DATE": 2,
+    "B-EVENT": 3,
+    "B-FAC": 4,
+    "B-GPE": 5,
+    "B-LANGUAGE": 6,
+    "B-LAW": 7,
+    "B-LOC": 8,
+    "B-MONEY": 9,
+    "B-NORP": 10,
+    "B-ORDINAL": 11,
+    "B-ORG": 12,
+    "B-PERCENT": 13,
+    "B-PERSON": 14,
+    "B-PRODUCT": 15,
+    "B-QUANTITY": 16,
+    "B-TIME": 17,
+    "B-WORK_OF_ART": 18,
+    "E-CARDINAL": 37,
+    "E-DATE": 38,
+    "E-EVENT": 39,
+    "E-FAC": 40,
+    "E-GPE": 41,
+    "E-LANGUAGE": 42,
+    "E-LAW": 43,
+    "E-LOC": 44,
+    "E-MONEY": 45,
+    "E-NORP": 46,
+    "E-ORDINAL": 47,
+    "E-ORG": 48,
+    "E-PERCENT": 49,
+    "E-PERSON": 50,
+    "E-PRODUCT": 51,
+    "E-QUANTITY": 52,
+    "E-TIME": 53,
+    "E-WORK_OF_ART": 54,
+    "I-CARDINAL": 19,
+    "I-DATE": 20,
+    "I-EVENT": 21,
+    "I-FAC": 22,
+    "I-GPE": 23,
+    "I-LANGUAGE": 24,
+    "I-LAW": 25,
+    "I-LOC": 26,
+    "I-MONEY": 27,
+    "I-NORP": 28,
+    "I-ORDINAL": 29,
+    "I-ORG": 30,
+    "I-PERCENT": 31,
+    "I-PERSON": 32,
+    "I-PRODUCT": 33,
+    "I-QUANTITY": 34,
+    "I-TIME": 35,
+    "I-WORK_OF_ART": 36,
+    "O": 0,
+    "S-CARDINAL": 55,
+    "S-DATE": 56,
+    "S-EVENT": 57,
+    "S-FAC": 58,
+    "S-GPE": 59,
+    "S-LANGUAGE": 60,
+    "S-LAW": 61,
+    "S-LOC": 62,
+    "S-MONEY": 63,
+    "S-NORP": 64,
+    "S-ORDINAL": 65,
+    "S-ORG": 66,
+    "S-PERCENT": 67,
+    "S-PERSON": 68,
+    "S-PRODUCT": 69,
+    "S-QUANTITY": 70,
+    "S-TIME": 71,
+    "S-WORK_OF_ART": 72
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 4,
+  "pad_token_id": 0,
+  "pooler_fc_size": 312,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
+  "tokenizer_class": "BertTokenizerFast",
+  "type_vocab_size": 2,
+  "vocab_size": 21128
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b42dfb2bf2752be7885dcbd0fbf8b2954bd07948645925df1ba569514c6f4a7e
+size 45891063

special_tokens_map.json

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

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": false, "do_basic_tokenize": true, "never_split": null, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "name_or_path": "bert-base-chinese"}