Upload 26 files

README.md

@@ -0,0 +1,84 @@
+---
+language:
+- zh
+license: apache-2.0
+tags:
+- whisper-event
+- generated_from_trainer
+datasets:
+- mozilla-foundation/common_voice_11_0
+metrics:
+- wer
+model-index:
+- name: Whisper medium nan-tw only char
+  results:
+  - task:
+      name: Automatic Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: mozilla-foundation/common_voice_11_0 nan-tw
+      type: mozilla-foundation/common_voice_11_0
+      config: nan-tw
+      split: test
+      args: nan-tw
+    metrics:
+    - name: Wer
+      type: wer
+      value: 45.2824427480916
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# Whisper medium nan-tw only char
+
+This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the mozilla-foundation/common_voice_11_0 nan-tw dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.9944
+- Wer: 45.2824
+- Cer: 45.3667
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 1e-05
+- train_batch_size: 2
+- eval_batch_size: 2
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_steps: 500
+- training_steps: 5000
+- mixed_precision_training: Native AMP
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Wer     | Cer     |
+|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|
+| 0.5832        | 1.04  | 1000 | 1.0634          | 56.3053 | 56.4745 |
+| 0.1467        | 2.08  | 2000 | 1.0407          | 50.9618 | 51.0112 |
+| 0.016         | 3.13  | 3000 | 1.0226          | 46.4427 | 46.5137 |
+| 0.0001        | 5.01  | 4000 | 0.9974          | 45.4656 | 45.6082 |
+| 0.0001        | 6.05  | 5000 | 0.9944          | 45.2824 | 45.3667 |
+
+
+### Framework versions
+
+- Transformers 4.27.0.dev0
+- Pytorch 1.13.1+cu117
+- Datasets 2.8.0
+- Tokenizers 0.13.2

added_tokens.json

@@ -0,0 +1,108 @@
+{
+  "<|af|>": 50327,
+  "<|am|>": 50334,
+  "<|ar|>": 50272,
+  "<|as|>": 50350,
+  "<|az|>": 50304,
+  "<|ba|>": 50355,
+  "<|be|>": 50330,
+  "<|bg|>": 50292,
+  "<|bn|>": 50302,
+  "<|bo|>": 50347,
+  "<|br|>": 50309,
+  "<|bs|>": 50315,
+  "<|ca|>": 50270,
+  "<|cs|>": 50283,
+  "<|cy|>": 50297,
+  "<|da|>": 50285,
+  "<|de|>": 50261,
+  "<|el|>": 50281,
+  "<|en|>": 50259,
+  "<|es|>": 50262,
+  "<|et|>": 50307,
+  "<|eu|>": 50310,
+  "<|fa|>": 50300,
+  "<|fi|>": 50277,
+  "<|fo|>": 50338,
+  "<|fr|>": 50265,
+  "<|gl|>": 50319,
+  "<|gu|>": 50333,
+  "<|haw|>": 50352,
+  "<|ha|>": 50354,
+  "<|he|>": 50279,
+  "<|hi|>": 50276,
+  "<|hr|>": 50291,
+  "<|ht|>": 50339,
+  "<|hu|>": 50286,
+  "<|hy|>": 50312,
+  "<|id|>": 50275,
+  "<|is|>": 50311,
+  "<|it|>": 50274,
+  "<|ja|>": 50266,
+  "<|jw|>": 50356,
+  "<|ka|>": 50329,
+  "<|kk|>": 50316,
+  "<|km|>": 50323,
+  "<|kn|>": 50306,
+  "<|ko|>": 50264,
+  "<|la|>": 50294,
+  "<|lb|>": 50345,
+  "<|ln|>": 50353,
+  "<|lo|>": 50336,
+  "<|lt|>": 50293,
+  "<|lv|>": 50301,
+  "<|mg|>": 50349,
+  "<|mi|>": 50295,
+  "<|mk|>": 50308,
+  "<|ml|>": 50296,
+  "<|mn|>": 50314,
+  "<|mr|>": 50320,
+  "<|ms|>": 50282,
+  "<|mt|>": 50343,
+  "<|my|>": 50346,
+  "<|ne|>": 50313,
+  "<|nl|>": 50271,
+  "<|nn|>": 50342,
+  "<|nocaptions|>": 50362,
+  "<|notimestamps|>": 50363,
+  "<|no|>": 50288,
+  "<|oc|>": 50328,
+  "<|pa|>": 50321,
+  "<|pl|>": 50269,
+  "<|ps|>": 50340,
+  "<|pt|>": 50267,
+  "<|ro|>": 50284,
+  "<|ru|>": 50263,
+  "<|sa|>": 50344,
+  "<|sd|>": 50332,
+  "<|si|>": 50322,
+  "<|sk|>": 50298,
+  "<|sl|>": 50305,
+  "<|sn|>": 50324,
+  "<|so|>": 50326,
+  "<|sq|>": 50317,
+  "<|sr|>": 50303,
+  "<|startoflm|>": 50360,
+  "<|startofprev|>": 50361,
+  "<|startoftranscript|>": 50258,
+  "<|su|>": 50357,
+  "<|sv|>": 50273,
+  "<|sw|>": 50318,
+  "<|ta|>": 50287,
+  "<|te|>": 50299,
+  "<|tg|>": 50331,
+  "<|th|>": 50289,
+  "<|tk|>": 50341,
+  "<|tl|>": 50348,
+  "<|transcribe|>": 50359,
+  "<|translate|>": 50358,
+  "<|tr|>": 50268,
+  "<|tt|>": 50351,
+  "<|uk|>": 50280,
+  "<|ur|>": 50290,
+  "<|uz|>": 50337,
+  "<|vi|>": 50278,
+  "<|yi|>": 50335,
+  "<|yo|>": 50325,
+  "<|zh|>": 50260
+}

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 6.05,
+    "eval_cer": 45.36673709628735,
+    "eval_loss": 0.994395911693573,
+    "eval_runtime": 226.1976,
+    "eval_samples_per_second": 4.359,
+    "eval_steps_per_second": 2.18,
+    "eval_wer": 45.2824427480916,
+    "train_loss": 0.3284327008752851,
+    "train_runtime": 3725.671,
+    "train_samples_per_second": 2.684,
+    "train_steps_per_second": 1.342
+}

config.json

@@ -0,0 +1,48 @@
+{
+  "_name_or_path": "openai/whisper-medium",
+  "activation_dropout": 0.0,
+  "activation_function": "gelu",
+  "apply_spec_augment": false,
+  "architectures": [
+    "WhisperForConditionalGeneration"
+  ],
+  "attention_dropout": 0.0,
+  "begin_suppress_tokens": [
+    220,
+    50257
+  ],
+  "bos_token_id": 50257,
+  "d_model": 1024,
+  "decoder_attention_heads": 16,
+  "decoder_ffn_dim": 4096,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 24,
+  "decoder_start_token_id": 50258,
+  "dropout": 0.0,
+  "encoder_attention_heads": 16,
+  "encoder_ffn_dim": 4096,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 24,
+  "eos_token_id": 50257,
+  "forced_decoder_ids": null,
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
+  "max_length": 448,
+  "max_source_positions": 1500,
+  "max_target_positions": 448,
+  "model_type": "whisper",
+  "num_hidden_layers": 24,
+  "num_mel_bins": 80,
+  "pad_token_id": 50257,
+  "scale_embedding": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.27.0.dev0",
+  "use_cache": true,
+  "vocab_size": 51865
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 6.05,
+    "eval_cer": 45.36673709628735,
+    "eval_loss": 0.994395911693573,
+    "eval_runtime": 226.1976,
+    "eval_samples_per_second": 4.359,
+    "eval_steps_per_second": 2.18,
+    "eval_wer": 45.2824427480916
+}

generation_config.json

@@ -0,0 +1,224 @@
+{
+  "begin_suppress_tokens": [
+    220,
+    50257
+  ],
+  "bos_token_id": 50257,
+  "decoder_start_token_id": 50258,
+  "eos_token_id": 50257,
+  "forced_decoder_ids": [
+    [
+      1,
+      null
+    ],
+    [
+      2,
+      50359
+    ],
+    [
+      3,
+      50363
+    ]
+  ],
+  "is_multilingual": true,
+  "lang_to_id": {
+    "<|af|>": 50327,
+    "<|am|>": 50334,
+    "<|ar|>": 50272,
+    "<|as|>": 50350,
+    "<|az|>": 50304,
+    "<|ba|>": 50355,
+    "<|be|>": 50330,
+    "<|bg|>": 50292,
+    "<|bn|>": 50302,
+    "<|bo|>": 50347,
+    "<|br|>": 50309,
+    "<|bs|>": 50315,
+    "<|ca|>": 50270,
+    "<|cs|>": 50283,
+    "<|cy|>": 50297,
+    "<|da|>": 50285,
+    "<|de|>": 50261,
+    "<|el|>": 50281,
+    "<|en|>": 50259,
+    "<|es|>": 50262,
+    "<|et|>": 50307,
+    "<|eu|>": 50310,
+    "<|fa|>": 50300,
+    "<|fi|>": 50277,
+    "<|fo|>": 50338,
+    "<|fr|>": 50265,
+    "<|gl|>": 50319,
+    "<|gu|>": 50333,
+    "<|haw|>": 50352,
+    "<|ha|>": 50354,
+    "<|he|>": 50279,
+    "<|hi|>": 50276,
+    "<|hr|>": 50291,
+    "<|ht|>": 50339,
+    "<|hu|>": 50286,
+    "<|hy|>": 50312,
+    "<|id|>": 50275,
+    "<|is|>": 50311,
+    "<|it|>": 50274,
+    "<|ja|>": 50266,
+    "<|jw|>": 50356,
+    "<|ka|>": 50329,
+    "<|kk|>": 50316,
+    "<|km|>": 50323,
+    "<|kn|>": 50306,
+    "<|ko|>": 50264,
+    "<|la|>": 50294,
+    "<|lb|>": 50345,
+    "<|ln|>": 50353,
+    "<|lo|>": 50336,
+    "<|lt|>": 50293,
+    "<|lv|>": 50301,
+    "<|mg|>": 50349,
+    "<|mi|>": 50295,
+    "<|mk|>": 50308,
+    "<|ml|>": 50296,
+    "<|mn|>": 50314,
+    "<|mr|>": 50320,
+    "<|ms|>": 50282,
+    "<|mt|>": 50343,
+    "<|my|>": 50346,
+    "<|ne|>": 50313,
+    "<|nl|>": 50271,
+    "<|nn|>": 50342,
+    "<|no|>": 50288,
+    "<|oc|>": 50328,
+    "<|pa|>": 50321,
+    "<|pl|>": 50269,
+    "<|ps|>": 50340,
+    "<|pt|>": 50267,
+    "<|ro|>": 50284,
+    "<|ru|>": 50263,
+    "<|sa|>": 50344,
+    "<|sd|>": 50332,
+    "<|si|>": 50322,
+    "<|sk|>": 50298,
+    "<|sl|>": 50305,
+    "<|sn|>": 50324,
+    "<|so|>": 50326,
+    "<|sq|>": 50317,
+    "<|sr|>": 50303,
+    "<|su|>": 50357,
+    "<|sv|>": 50273,
+    "<|sw|>": 50318,
+    "<|ta|>": 50287,
+    "<|te|>": 50299,
+    "<|tg|>": 50331,
+    "<|th|>": 50289,
+    "<|tk|>": 50341,
+    "<|tl|>": 50348,
+    "<|tr|>": 50268,
+    "<|tt|>": 50351,
+    "<|uk|>": 50280,
+    "<|ur|>": 50290,
+    "<|uz|>": 50337,
+    "<|vi|>": 50278,
+    "<|yi|>": 50335,
+    "<|yo|>": 50325,
+    "<|zh|>": 50260
+  },
+  "max_initial_timestamp_index": 1,
+  "max_length": 448,
+  "no_timestamps_token_id": 50363,
+  "pad_token_id": 50257,
+  "suppress_tokens": [
+    1,
+    2,
+    7,
+    8,
+    9,
+    10,
+    14,
+    25,
+    26,
+    27,
+    28,
+    29,
+    31,
+    58,
+    59,
+    60,
+    61,
+    62,
+    63,
+    90,
+    91,
+    92,
+    93,
+    359,
+    503,
+    522,
+    542,
+    873,
+    893,
+    902,
+    918,
+    922,
+    931,
+    1350,
+    1853,
+    1982,
+    2460,
+    2627,
+    3246,
+    3253,
+    3268,
+    3536,
+    3846,
+    3961,
+    4183,
+    4667,
+    6585,
+    6647,
+    7273,
+    9061,
+    9383,
+    10428,
+    10929,
+    11938,
+    12033,
+    12331,
+    12562,
+    13793,
+    14157,
+    14635,
+    15265,
+    15618,
+    16553,
+    16604,
+    18362,
+    18956,
+    20075,
+    21675,
+    22520,
+    26130,
+    26161,
+    26435,
+    28279,
+    29464,
+    31650,
+    32302,
+    32470,
+    36865,
+    42863,
+    47425,
+    49870,
+    50254,
+    50258,
+    50358,
+    50359,
+    50360,
+    50361,
+    50362
+  ],
+  "task_to_id": {
+    "transcribe": 50359,
+    "translate": 50358
+  },
+  "transformers_version": "4.27.0.dev0"
+}

label.txt

@@ -0,0 +1 @@
+label_str=['心 理 性 別 ', '火 燒 尻 川 ', '草 仔 色 ', '十 三 行 ', '水 里 鄉 ', '這 搭 有 一 个 大 商 場 ', '聯 合 國 ', '菜 頭 籠 仔 ', '辣 椒 ', '肉 筋 ', '新 興 區 ', '癩 疒 哥 病 ', '覕 冬 ', '摧 仔 ', '西 藥 房 ', '心 碎 ', '燒 包 仔 ', '通 敨 ', '皇 軍 ', '濁 水 溪 出 代 誌 ', '粗 細 ', '哀 哼 ', '九 年 一 貫 ', '精 靈 ', '三 寶 ', '速 速 ', '坐 疶 咧 等 ', '儑 目 袂 曉 看 懸 低 ', '圳 後 ', '兩 个 人 已 經 講 和 矣 ', '足 媠 ', '細 空 喙 ', '掠 著 根 頭 ', '茨 城 縣 ', '概 率 ', '草 人 坑 ', '祝 你 生 日 快 樂 ', '坑 崁 ', '出 水 ', '咇 怦 喘 ', '夯 枷 ', '外 勞 ', '臺 灣 人 ', '落 湳 ', '苑 裡 坑 ', '坑 仔 內 ', '起 慼 到 地 ', '雨 晴 ', '紅 蟲 ', '大 湖 ', '誻 誻 叫 ', '柴 球 ', '大 胃 王 ', '大 稻 埕 ', '偷 壘 ', '鬆 餅 ', '喙 罨 囊 ', '龍 文 ', '時 局 ', '摔 死 ', '煞 尾 ', '無 話 講 茭 荖 ', '下 暗 飯 ', '一 籠 ', '一 定 考 袂 牢 ', '看 一 个 影 生 一 个 囝 ', '倒 踅 ', '筆 型 ', '歸 正 修 善 ', '滯 到 ', '變 天 ', '一 工 到 暗 ', '雲 吞 ', '有 幾 項 菜 ', '炎 舞 ', '鹿 角 龜 ', '過 路 線 ', '惡 馬 惡 人 騎 ', '粗 俗 ', '這 个 偌 濟 錢 ', '七 仔 ', '看 甲 凊 凊 楚 楚 ', '戇 神 ', '發 生 ', '病 囡 仔 ', '心 花 當 開 ', '番 鼎 仔 ', '老 硞 硞 ', '大 傢 伙 ', '名 聲 真 敨 ', '攕 烘 肉 ', '滑 ', '龜 蛇 ', '水 雞 皮 仔 ', '攑 硬 篙 ', '抾 無 三 門 墓 著 想 欲 做 土 公 ', '紅 圓 仔 ', '臺 北 橋 ', '雲 林 縣 ', '臺 中 港 ', '水 里 ', '臺 南 市 ', '七 堵 ', '做 功 德 ', '大 村 鄉 ', '小 喇 叭 ', '紹 ', '漢 草 真 好 ', '田 嬰 ', '按 呢 生 ', '海 尾 ', '無 米 閣 拄 著 閏 月 ', '枝 香 小 細 ', '噗 ', '頭 尾 ', '五 短 生 張 ', '正 常 ', '套 牢 咧 ', '動 物 性 膠 質 ', '雪 文 精 ', '悠 遊 卡 ', '你 攏 無 歇 睏 喔 ', '埤 頭 鄉 ', '信 義 大 安 路 口 ', '三 三 八 八 ', '你 話 毋 通 烏 白 講 ', '麥 寮 鄉 ', '你 莫 定 定 挖 肚 臍 ', '新 安 路 ', '金 錢 毋 是 萬 能 的 ', '菠 菱 仔 菜 ', '富 寮 里 ', '你 是 啥 人 ', '反 烏 ', '新 街 ', '合 銅 ', '疼 疼 ', '番 麥 殼 ', '天 氣 冷 矣 你 就 愛 加 疊 一 領 衫 ', '我 到 底 看 著 啥 物 ', '手 扞 仔 ', '姑 不 二 三 將 ', '品 格 ', '透 種 仔 ', '漚 腳 數 ', '隔 音 壁 ', '拍 通 透 ', '開 花 滿 天 芳 結 子 才 驚 人 ', '東 北 季 風 ', '無 人 機 ', '做 牛 無 惜 力 ', '淨 空 排 放 ', '這 藥 仔 抹 落 去 真 緊 就 會 退 癀 矣 ', '毛 蓑 仔 ', '抉 喙 䫌 ', '陽 明 山 ', '咖 啡 館 ', '囡 仔 色 ', '佇 墜 樓 前 一 點 鐘 ', '伊 規 心 欲 考 大 學 ', '東 湖 國 中 ', '中 原 公 園 ', '半 路 店 ', '檳 榔 坑 ', '東 石 鄉 ', '摒 掃 的 ', '無 中 心 化 ', '肉 粽 角 ', '編 譯 器 ', '厝 邊 頭 尾 ', '桃 源 街 ', '仁 愛 杭 州 路 口 ', '千 葉 ', '七 色 圖 ', '大 吉 大 利 ', '尪 仔 頭 鏡 ', '徐 ', '咱 做 代 誌 的 時 陣 ', '包 水 餃 ', '歹 鳥 毋 知 飛 歹 柴 破 袂 開 ', '分 裝 場 ', '健 康 中 心 ', '印 度 尼 西 亞 ', '新 里 族 ', '獅 仔 頭 ', '大 度 路 ', '市 仔 ', '露 營 區 ', '學 姐 ', '枋 南 線 ', '虎 尾 鎮 ', '瀉 腹 肚 ', '恆 春 鎮 ', '小 血 球 ', '破 格 ', '我 衫 攏 穿 上 大 領 ', '溜 落 來 ', '半 條 命 ', '生 活 習 慣 ', '電 影 間 ', '菜 脯 卵 ', '塗 塗 塗 ', '鐵 甲 牛 ', '風 火 頭 ', '現 流 仔 ', '白 了 工 ', '空 港 ', '較 大 ', '公 正 橋 ', '巴 結 ', '果 汁 ', '有 䆀 無 一 好 ', '這 時 ', '竹 北 ', '下 早 ', '較 早 睏 ', '六 六 大 順 ', '公 視 ', '趕 趕 緊 緊 ', '帕 仔 ', '藍 瘦 香 菇 ', '山 貓 ', '大 頭 狗 母 ', '土 板 仁 豆 ', '咧 ', '內 緣 ', '阮 會 曉 唸 ', '茶 末 ', '真 有 喙 水 ', '魔 導 士 ', '先 承 認 你 就 是 恁 朋 友 ', '青 山 王 ', '奧 克 蘭 ', '金 東 ', '便 若 ', '陳 奕 齊 ', '九 尾 金 ', '貓 空 ', '過 tang跤 ', '頭 汴 坑 溪 ', '無 定 著 ', '控 機 仔 ', '趨 湧 ', '中 洲 ', '青 田 街 ', '同 房 ', '鼢 底 的 ', '多 元 性 別 ', '舊 厝 ', '貓 潲 ', '福 連 國 小 ', '點 撇 ', '玜 玳 ', '欲 起 鼓 矣 ', '納 豆 ', '中 環 ', '你 死 我 活 ', '踏 著 歹 地 步 ', '曲 去 ', '怙 喙 講 个 無 準 算 ', '交 力 坪 ', '愈 來 愈 嚴 重 ', '一 个 一 个 ', '淡 江 大 橋 ', '上 大 代 死 煞 ', '檨 仔 葉 公 ', '胳 胴 跤 ', '郁 永 河 ', '講 啥 物 芋 仔 番 薯 ', '霓 ', '青 盲 雞 啄 著 蟲 ', '搣 屎 毋 知 臭 ', '斗 南 ', '孤 笑 無 嗣 ', '水 鬼 仔 ', '門 將 ', '手 控 ', '心 胞 膜 炎 ', '敢 食 汽 油 ', '阿 卜 蛇 ', '機 動 戰 士 ', '便 當 ', '敧 一 爿 ', '敲 仙 古 ', '所 羅 門 群 島 ', '終 界 ', '我 愛 台 語 ', '溝 頂 ', '墊 定 ', '詳 解 ', '火 燒 ', '絡 繹 不 ��� ', '尻 川 門 ', '甌 仔 疊 碟 仔 ', '黃 巾 甕 仔 ', '汕 ', '生 甲 袂 䆀 ', '狗 母 酥 ', '倚 桌 仔 ', '中 央 大 學 ', '石 門 水 庫 ', '亂 使 想 ', '百 貨 店 ', '酸 素 矸 ', '蛤 kai', '洗 衫 仔 枋 ', '天 台 ', '數 位 pay', '漚 缺 ', '鈍 市 貨 ', '偷 走 壘 ', '夭 壽 鬼 ', '好 所 在 ', '軟 市 ', '徹 理 ', '現 挽 ', '叫 電 話 ', '伊 生 做 遮 爾 臭 老 ', '魔 神 仔 咒 讖 ', '一 點 點 仔 ', '兩 光 ', '契 查 某 囝 ', '提 著 ', '傷 心 酒 店 ', '過 失 致 死 ', '較 拚 咧 ', '透 身 大 汗 ', '碌 硞 馬 仔 ', '兄 弟 姊 妹 著 愛 仝 心 ', '家 伙 仔 ', '垃 圾 物 仔 ', '狐 狸 貂 ', '石 斑 ', '虎 魚 ', '落 雨 了 後 塗 跤 變 甲 澹 漉 漉 ', '較 慘 ', '基 隆 市 ', '你 欲 行 對 佗 位 去 ', '敦 化 和 平 路 口 ', '港 仔 喙 ', '阮 囝 今 年 二 十 出 頭 矣 ', '新 港 飴 ', '哀 父 叫 母 ', '一 份 ', '毛 毛 仔 雨 ', '磐 安 ', '牛 籠 ', '袂 看 得 ', '立 功 立 德 ', '龜 山 島 ', '過 晝 ', '捒 做 堆 ', '梅 仔 跤 ', '包 帶 ', '番 地 ', '三 重 ', '林 子 口 ', '筆 電 ', '操 刀 ', '揣 轉 來 ', '指 中 指 ', '珠 仔 台 ', '新 廍 ', '相 帶 ', '頂 塊 廖 ', '竹 南 ', '你 共 包 仔 提 去 熁 燒 一 下 ', '平 方 根 ', '頂 公 館 ', '散 仙 ', '演 藝 界 ', '繪 本 ', '毋 情 毋 願 ', '較 緊 ', '有 也 好 無 也 好 ', '操 心 ', '摃 槌 仔 龍 ', '怪 奇 古 ', '國 破 家 亡 ', '新 羅 ', '貴 氣 ', '沓 沓 仔 ', '拍 薄 ', '物 件 若 予 伊 提 過 手 ', '輾 仔 鞋 ', '異 界 ', '鼠 尾 風 ', '湖 內 ', '提 走 ', '無 打 無 啥 ', '我 欲 準 備 落 山 矣 ', '盜 用 ', '聽 無 ', '彼 个 頭 家 娘 對 人 客 誠 好 禮 ', '大 細 輦 ', '無 歇 ', '無 疑 悟 你 會 出 國 讀 冊 ', '石 園 ', '嘉 義 ', '後 寮 ', '我 攏 有 看 著 ', '劍 南 路 ', '歹 天 ', '三 跤 步 一 坎 店 ', '臘 肉 ', '舊 里 族 ', '下 晡 點 心 ', '冷 管 ', '無 及 格 ', '公 允 ', '十 喙 九 尻 川 ', '投 影 片 ', '納 錢 ', '墓 坑 鳥 ', '活 餌 桶 仔 ', '激 屎 面 ', '袂 對 同 ', '紡 織 機 ', '木 質 部 ', '落 口 講 出 來 ', '投 丁 ', '棚 頂 做 甲 流 汗 棚 跤 嫌 甲 流 瀾 ', '僫 鬥 陣 ', '拍 死 板 ', '捅 鼻 ', '搭 峇 ', '簡 述 ', '一 个 接 一 个 ', '緊 來 旋 ', '閒 仔 話 ', '拍 爽 ', '德 拉 瓦 ', '總 舖 ', '三 日 不 見 久 溜 溜 ', '合 掛 ', '手 神 重 ', '電 子 ', '暗 空 ', '瘦 食 ', '塑 膠 車 ', '要 意 ', '燒 雞 胿 ', '吊 猴 ', '掠 猴 ', '暗 頭 仔 ', '濟 囝 濟 擘 腹 濟 新 婦 濟 體 剔 ', '交 球 ', '無 拄 好 ', '搶 頭 標 ', '金 環 失 日 ', '觀 音 大 士 ', '在 準 ', '欲 暗 矣 ', '嘔 落 ', '蛣 蛚 ', '一 目 仔 ', '提 款 片 ', '無 你 个 代 ', '天 公 祖 ', '串 講 ', '離 題 ', '馬 鈴 薯 條 ', '做 塗 水 的 ', '人 情 味 ', '劉 銘 傳 ', '肝 包 油 ', '糖 醋 ', '二 戰 ', '拍 鐵 師 ', '慢 一 跤 步 ', '提 批 的 ', '下 禮 拜 ', '死 亡 之 握 ', '你 毋 通 囥 佇 心 肝 內 ', '臭 龜 仔 ', '趁 一 空 食 三 冬 ', '我 欲 學 臺 語 ', '埋 ', '馬 卡 道 ', '箍 喙 罨 ', '貴 州 省 ', '敦 化 北 路 ', '你 看 起 來 是 遮 爾 仔 媠 ', '精 个 食 戇 戇 个 食 天 公 ', '轉 去 ', '下 晡 市 ', '幾 个 ', '邦 長 ', '港 墘 ', '字 條 仔 ', '無 去 矣 ', '言 行 錄 ', '買 空 ', '六 點 ', '死 忠 ', '石 母 奶 ', '耳 仔 機 ', '貢 寮 區 ', '含 血 噴 天 ', '大 好 額 人 ', '行 政 院 ', '食 市 ', '九 塊 厝 ', '冷 清 ', '月 光 ', '種 喙 齒 ', '街 友 ', '天 公 星 ', '縣 ', '愛 臺 語 ', '我 強 欲 袂 赴 矣 ', '添 飯 ', '有 連 線 ', '昶 ', '中 國 話 ', '攑 香 綴 拜 ', '四 跤 草 ', '疼 風 ', '觀 覽 ', '吉 祥 卵 ', '擲 掉 ', '打 狗 山 ', '西 瓜 倚 大 爿 ', '有 這 款 的 代 誌 我 攏 毋 知 ', '磨 豆 仔 機 ', '國 姓 爺 ', '林 口 區 ', '潘 ', '可 算 名 詞 ', '代 誌 大 碗 ', '半 信 半 疑 ', '恬 稚 恬 稚 討 客 兄 唯 是 ', '遛 光 光 ', '扲 糖 無 洗 手 的 朋 友 ', '烏 日 ', '魏 風 ', '膵 臟 ', '犯 險 ', '連 半 个 人 影 嘛 無 ', '吊 褲 ', '塗 水 工 ', '霧 嗄 嗄 ', '飛 行 船 ', '欲 完 矣 ', '運 將 ', '恁 老 母 拄 咧 念 你 ', '伊 有 誠 濟 內 孫 ', '港 仔 喙 國 小 ', '橐 個 束 個 ', '落 尾 ', '平 枋 電 腦 ', '毛 毛 仔 雨 落 久 塗 塗 嘛 會 澹 ', '冷 滾 水 ', '一 个 紅 龜 按 一 个 位 ', '鹹 酥 雞 ', '雞 卵 糕 仔 ', '溼 汗 ', '諸 島 ', '聖 誕 老 阿 公 ', '天 頂 白 茫 茫 ', '飯 煎 ', '男 權 ', '毋 認 輸 ', '餘 溫 ', '包 山 包 海 ', '踮 空 龜 ', '洗 衫 精 ', '迒 境 電 商 ', '開 喙 蚶 粒 粒 臭 ', '焦 水 期 ', '老 龜 精 ', '塑 膠 橐 仔 ', '蘇 維 埃 聯 盟 ', '臭 水 ', '束 頷 頸 仔 ', '飛 行 機 ', '平 溪 ', '唯 恐 天 下 不 亂 ', '錢 票 ', '瓜 仔 鬚 ', '羊 角 三 劍 ', '半 腰 ', '辭 彙 ', '衝 湧 ', '敢 有 問 題 ', '食 涼 ', '畫 蛇 添 足 ', '做 塗 水 ', '怪 奇 ', '板 金 ', '交 關 ', '海 牙 ', '中 山 路 ', '番 仔 狗 ', '粉 圓 仔 ', '龜 龜 毛 毛 ', '毋 捌 字 看 告 示 ', '表 單 ', '莒 光 號 ', '海 豬 仔 ', '猴 頭 果 ', '翕 新 娘 相 ', '面 書 ', '茶 葉 蛋 ', '花 枝 招 展 ', '食 人 一 口 還 人 一 斗 ', '鐵 彈 ', '東 白 仔 ', '年 兜 ', '佛 心 投 資 者 ', '人 講 綴 人 講 ', '金 剛 戰 士 ', '做 肉 餅 ', '相 駁 喙 ', '鸚 哥 魚 ', '扁 桃 泉 ', '一 筆 錢 ', '果 菜 ', '柑 仔 蜜 醬 ', '一 枝 草 一 點 露 ', '早 時 ', '大 頭 母 人 ', '包 的 ', '頂 禮 拜 ', '尿 苴 仔 ', '人 驚 鬼 鬼 驚 人 ', '無 記 持 症 ', '日 落 山 ', '偏 鄉 ', '拭 塗 跤 ', '袂 死 心 ', '蓋 濟 ', '左 右 鄰 ', '有 聽 著 無 ', '大 圓 埕 ', '斗 籠 仔 本 ', '摃 寇 ', '野 蓮 ', '勥 查 某 ', '暗 頓 食 啥 ', '草 其 略 仔 ', '暢 快 ', '小 包 ', '薄 餅 ', '趒 ', '暖 帽 ', '無 精 彩 ', '內 家 ', '鬍 鬚 ', '燒 仙 草 ', '痀 崙 ', '吳 剛 剉 桂 ', '沙 灘 ', '人 聲 喊 喝 ', '天 公 無 帶 著 ', '望 安 鄉 ', '食 風 ', '分 子 ', '批 車 ', '大 舌 ', '公 館 路 ', '碧 潭 路 ', '臺 北 歐 洲 學 校 小 學 部 ', '充 耳 不 聞 ', '四 份 仔 ', '頭 崁 的 ', '南 機 場 ', '永 綏 街 ', '朴 子 溪 ', '池 上 ', '跤 車 ', '管 待 ', '苦 袂 當 ', '陳 滄 江 ', '缺 囝 化 ', '硨 磲 ', '共 人 唌 ', '頂 塊 黃 ', '潮 州 街 口 ', '社 群 ', '鄉 長 厝 ', '豐 樂 公 園 ', '金 湖 鎮 ', '忠 孝 復 興 ', '民 族 區 運 路 口 ', '松 信 路 ', '糋 麭 粉 ', '鄞 州 ', '台 北 聯 絡 線 ', '又 閣 咧 咳 嗽 ', '伊 加 我 一 輩 ', '霜 仔 ', '魚 池 鄉 ', '囝 甘 仔 囝 ', '四 跤 水 仙 ', '暖 暖 區 ', '用 齒 抿 仔 洗 喙 時 愛 先 捋 齒 膏 ', '新 社 ', '母 身 若 顧 予 伊 好 ', '實 喉 ', '埔 尾 ', '步 路 ', '冰 糖 ', '復 興 橋 ', '阿 拉 ', '繏 絚 ', '員 林 市 ', '蝦 糋 ', '戇 代 誌 ', '年 拄 年 ', '景 興 路 ', '海 結 仔 頭 ', '一 過 半 過 ', '實 聯 制 ', '後 擴 ', '瑞 芳 ', '蘭 嶼 貓 頭 鳥 ', '貓 羅 溪 ', '子 宮 喙 癌 ', '果 子 宅 ', '假 說 ', '不 在 ', '燕 巢 區 ', '上 海 毛 蟹 ', '意 式 麵 ', '捲 風 ', '話 語 之 中 ', '皮 蛋 精 肉 糜 ', '紅 豆 仔 麭 ', '頂 回 ', '相 見 ', '三 兩 人 講 四 斤 話 ', '梘 尾 ', '質 子 ', '兩 箍 ', '真 珠 看 做 鳥 鼠 屎 ', '天 晴 ', '角 運 動 量 ', '莫 問 足 恐 怖 ', '鼻 隔 間 ', '叨 撚 ', '止 疼 藥 ', '恙 蟲 病 ', '火 成 岩 ', '許 崑 源 ', '土 象 ', '豬 頭 癀 ', '禁 喙 ', '昨 昏 ', '卡 通 ', '番 薯 片 ', '砰 去 ', '卵 珠 ', '歡 喜 甘 願 ', '大 海 毋 驚 大 水 ', '橫 的 ', '跤 手 賤 ', '旭 ', '仙 巴 掌 ', '徛 桌 櫃 ', '查 畝 營 ', '番 薯 球 ', '芳 料 草 ', '大 火 ', '夥 計 某 ', '英 國 話 ', '明 知 故 問 ', '老 街 ', '看 有 ', '燃 柴 添 火 著 ', '藝 文 活 動 ', '臺 羅 字 ', '利 ', '金 煌 檨 ', '死 袂 見 笑 ', '嚇 止 ', '加 法 ', '馬 鈴 薯 片 ', '日 頭 曝 尻 川 矣 閣 毋 起 來 ', '英 語 課 本 ', '凍 水 機 ', '大 肚 胿 仔 ', '卑 南 ', '袂 當 干 焦 我 看 著 ', '牛 睏 山 ', '代 誌 好 勢 矣 未 ', '媠 氣 ', '景 文 街 ', '咒 天 咒 地 ', '今 年 是 一 个 好 年 冬 ', '事 假 ', '烏 山 頭 ', '三 講 四 毋 著 ', '語 素 文 字 ', '反 差 ', '大 日 本 帝 國 ', '迒 界 網 商 ', '嚨 喉 ', '褪 躘 ', '莫 假 ', '緊 去 做 代 誌 較 要 緊 ', '你 講 啥 ', '頸 椎 ', '違 扴 ', '身 體 健 康 ', '跤 手 十 足 ', '電 磁 學 ', '繫 年 ', '大 鑼 大 鼓 ', '伊 是 一 个 大 美 人 ', '解 溶 ', '磅 去 ', '你 陪 我 去 彼 爿 好 無 ', '網 路 用 甲 牢 咧 ', '酒 駛 ', '孤 單 一 身 ', '家 治 ', '死 忠 迷 眾 ', '浮 洲 仔 ', '孤 行 獨 市 ', '日 冕 ', '注 目 ', '擉 仔 ', '燒 狗 ', '名 喙 ', '創 辦 人 ', '水 燦 林 ', '飼 奶 動 物 ', '目 睭 擘 金 ', '抾 囡 仔 ', '我 予 滾 水 燙 一 下 膨 疱 足 疼 的 ', '象 桮 ', '有 的 人 ', '瓜 月 ', '三 字 經 ', '飯 焦 ', '天 車 ', '梅 毒 ', '目 仔 久 ', '耙 形 ', '便 物 ', '鹽 酸 草 ', '番 仔 番 薯 ', '每 一 擺 來 你 攏 煮 好 料 的 ', '來 喔 緊 來 看 喔 ', '番 界 ', '總 書 記 ', '破 柴 ', '列 位 ', '鳳 山 ', '劉 厝 ', '凱 旋 武 昌 ', '金 美 國 小 ', '竹 北 市 ', '橋 頭 區 ', '南 州 ', '和 美 鎮 ', '中 壢 ', '西 區 ', '國 泰 ', '水 長 流 ', '中 央 路 ', '南 福 德 ', '金 華 大 廈 ', '獅 甲 ', '士 林 區 ', '紙 票 變 厚 ', '中 山 橋 ', '彰 化 銀 行 ', '黃 國 昌 ', '翠 峰 橋 ', '花 仔 菜 ', '中 正 國 小 ', '按 呢 就 好 ', '老 去 矣 ', '車 埕 ', '南 陽 街 ', '匈 奴 ', '松 壽 路 口 ', '中 坑 ', '不 速 之 客 ', '瑞 源 ', '後 驛 ', '大 稻 埕 ', '鳥 仔 弓 ', '再 敗 ', '新 營 區 ', '正 義 郵 局 ', '寧 夏 路 ', '動 漫 ', '大 和 園 ', '仁 愛 鄉 ', '快 龍 ', '範 ', '西 面 ', '食 鹼 ', '吊 懸 低 ', '先 生 娘 ', '拍 鐵 寮 ', '露 營 區 ', '韻 尾 ', '外 雙 溪 ', '信 義 區 ', '恁 阿 母 敢 知 影 你 佇 遮 發 廢 文 ', '敢 按 呢 ', '厚 操 煩 ', '望 星 橋 ', '姊 夫 ', '十 二 甲 ', '文 昌 橋 ', '有 要 緊 無 ', '林 內 鄉 ', '紅 目 石 獅 ', '囡 仔 人 嘛 看 有 ', '聽 帶 ', '歕 鼓 吹 ', '歹 喙 斗 ', '查 某 營 ', '滷 菜 頭 ', '跳 過 ', '鴨 母 拖 秤 錘 ', '不 貪 不 取 ', '契 兄 ', '白 墨 粉 ', '天 安 門 廣 場 ', '投 文 ', '隱 名 化 ', '粗 花 ', '雞 籠 蜂 ', '暗 眠 摸 山 貓 ', '袂 曉 講 話 ', '搖 床 ', '順 序 ', '捎 攏 無 ', '企 業 家 ', '梯 度 ', '足 久 無 看 見 ', '起 雞 母 皮 ', '台 東 ', '不 見 天 ', '內 華 達 ', '釣 蝦 仔 場 ', '罟 寮 仔 ', '五 支 指 頭 仔 咬 起 來 逐 支 嘛 疼 ', '羊 毛 揻 ', '厝 尾 頂 ', '扞 事 人 ', '大 箍 ', '較 低 ', '客 服 ', '鴨 母 泅 ', '蚵 仔 菇 ', '海 口 ', '阿 里 不 達 ', '反 桌 ', '截 ', '屎 啦 ', '臭 屎 星 ', '摸 王 爺 尻 川 ', '切 腹 ', '雄 三 飛 彈 ', '一 切 攏 是 假 的 ', '小 山 崙 仔 ', '剪 仔 龜 ', '青 盲 仔 目 鏡 ', '親 愛 的 ', '在 地 大 漢 ', '欲 去 佗 位 ', '崁 頭 鴨 ', '冕 旒 ', '跩 著 ', '雞 仔 腸 鳥 仔 肚 ', '門 擋 仔 ', '菱 角 龍 ', '狗 頭 芙 蓉 ', '癱 瘓 ', '我 欲 去 便 所 ', '彼 當 時 ', '鼻 淚 管 ', '代 誌 毋 好 ', '浸 透 ', '瓦 斯 彈 ', '材 料 科 學 ', '甕 底 水 雞 ', '匯 錢 ', '睏 袂 去 ']

normalizer.json

@@ -0,0 +1,1742 @@
+{
+  "accessorise": "accessorize",
+  "accessorised": "accessorized",
+  "accessorises": "accessorizes",
+  "accessorising": "accessorizing",
+  "acclimatisation": "acclimatization",
+  "acclimatise": "acclimatize",
+  "acclimatised": "acclimatized",
+  "acclimatises": "acclimatizes",
+  "acclimatising": "acclimatizing",
+  "accoutrements": "accouterments",
+  "aeon": "eon",
+  "aeons": "eons",
+  "aerogramme": "aerogram",
+  "aerogrammes": "aerograms",
+  "aeroplane": "airplane",
+  "aeroplanes": "airplanes",
+  "aesthete": "esthete",
+  "aesthetes": "esthetes",
+  "aesthetic": "esthetic",
+  "aesthetically": "esthetically",
+  "aesthetics": "esthetics",
+  "aetiology": "etiology",
+  "ageing": "aging",
+  "aggrandisement": "aggrandizement",
+  "agonise": "agonize",
+  "agonised": "agonized",
+  "agonises": "agonizes",
+  "agonising": "agonizing",
+  "agonisingly": "agonizingly",
+  "almanack": "almanac",
+  "almanacks": "almanacs",
+  "aluminium": "aluminum",
+  "amortisable": "amortizable",
+  "amortisation": "amortization",
+  "amortisations": "amortizations",
+  "amortise": "amortize",
+  "amortised": "amortized",
+  "amortises": "amortizes",
+  "amortising": "amortizing",
+  "amphitheatre": "amphitheater",
+  "amphitheatres": "amphitheaters",
+  "anaemia": "anemia",
+  "anaemic": "anemic",
+  "anaesthesia": "anesthesia",
+  "anaesthetic": "anesthetic",
+  "anaesthetics": "anesthetics",
+  "anaesthetise": "anesthetize",
+  "anaesthetised": "anesthetized",
+  "anaesthetises": "anesthetizes",
+  "anaesthetising": "anesthetizing",
+  "anaesthetist": "anesthetist",
+  "anaesthetists": "anesthetists",
+  "anaesthetize": "anesthetize",
+  "anaesthetized": "anesthetized",
+  "anaesthetizes": "anesthetizes",
+  "anaesthetizing": "anesthetizing",
+  "analogue": "analog",
+  "analogues": "analogs",
+  "analyse": "analyze",
+  "analysed": "analyzed",
+  "analyses": "analyzes",
+  "analysing": "analyzing",
+  "anglicise": "anglicize",
+  "anglicised": "anglicized",
+  "anglicises": "anglicizes",
+  "anglicising": "anglicizing",
+  "annualised": "annualized",
+  "antagonise": "antagonize",
+  "antagonised": "antagonized",
+  "antagonises": "antagonizes",
+  "antagonising": "antagonizing",
+  "apologise": "apologize",
+  "apologised": "apologized",
+  "apologises": "apologizes",
+  "apologising": "apologizing",
+  "appal": "appall",
+  "appals": "appalls",
+  "appetiser": "appetizer",
+  "appetisers": "appetizers",
+  "appetising": "appetizing",
+  "appetisingly": "appetizingly",
+  "arbour": "arbor",
+  "arbours": "arbors",
+  "archaeologically": "archeologically",
+  "archaeologist": "archeologist",
+  "archaeologists": "archeologists",
+  "archaeology": "archeology</span>",
+  "archeological": "archaeological",
+  "ardour": "ardor",
+  "armour": "armor",
+  "armoured": "armored",
+  "armourer": "armorer",
+  "armourers": "armorers",
+  "armouries": "armories",
+  "armoury": "armory",
+  "artefact": "artifact",
+  "artefacts": "artifacts",
+  "authorise": "authorize",
+  "authorised": "authorized",
+  "authorises": "authorizes",
+  "authorising": "authorizing",
+  "axe": "ax",
+  "backpedalled": "backpedaled",
+  "backpedalling": "backpedaling",
+  "bannister": "banister",
+  "bannisters": "banisters",
+  "baptise": "baptize",
+  "baptised": "baptized",
+  "baptises": "baptizes",
+  "baptising": "baptizing",
+  "bastardise": "bastardize",
+  "bastardised": "bastardized",
+  "bastardises": "bastardizes",
+  "bastardising": "bastardizing",
+  "battleax": "battleaxe",
+  "baulk": "balk",
+  "baulked": "balked",
+  "baulking": "balking",
+  "baulks": "balks",
+  "bedevilled": "bedeviled",
+  "bedevilling": "bedeviling",
+  "behaviour": "behavior",
+  "behavioural": "behavioral",
+  "behaviourism": "behaviorism",
+  "behaviourist": "behaviorist",
+  "behaviourists": "behaviorists",
+  "behaviours": "behaviors",
+  "behove": "behoove",
+  "behoved": "behooved",
+  "behoves": "behooves",
+  "bejewelled": "bejeweled",
+  "belabour": "belabor",
+  "belaboured": "belabored",
+  "belabouring": "belaboring",
+  "belabours": "belabors",
+  "bevelled": "beveled",
+  "bevvies": "bevies",
+  "bevvy": "bevy",
+  "biassed": "biased",
+  "biassing": "biasing",
+  "bingeing": "binging",
+  "bougainvillaea": "bougainvillea",
+  "bougainvillaeas": "bougainvilleas",
+  "bowdlerise": "bowdlerize",
+  "bowdlerised": "bowdlerized",
+  "bowdlerises": "bowdlerizes",
+  "bowdlerising": "bowdlerizing",
+  "breathalyse": "breathalyze",
+  "breathalysed": "breathalyzed",
+  "breathalyser": "breathalyzer",
+  "breathalysers": "breathalyzers",
+  "breathalyses": "breathalyzes",
+  "breathalysing": "breathalyzing",
+  "brutalise": "brutalize",
+  "brutalised": "brutalized",
+  "brutalises": "brutalizes",
+  "brutalising": "brutalizing",
+  "busses": "buses",
+  "bussing": "busing",
+  "caesarean": "cesarean",
+  "caesareans": "cesareans",
+  "calibre": "caliber",
+  "calibres": "calibers",
+  "calliper": "caliper",
+  "callipers": "calipers",
+  "callisthenics": "calisthenics",
+  "canalise": "canalize",
+  "canalised": "canalized",
+  "canalises": "canalizes",
+  "canalising": "canalizing",
+  "cancelation": "cancellation",
+  "cancelations": "cancellations",
+  "cancelled": "canceled",
+  "cancelling": "canceling",
+  "candour": "candor",
+  "cannibalise": "cannibalize",
+  "cannibalised": "cannibalized",
+  "cannibalises": "cannibalizes",
+  "cannibalising": "cannibalizing",
+  "canonise": "canonize",
+  "canonised": "canonized",
+  "canonises": "canonizes",
+  "canonising": "canonizing",
+  "capitalise": "capitalize",
+  "capitalised": "capitalized",
+  "capitalises": "capitalizes",
+  "capitalising": "capitalizing",
+  "caramelise": "caramelize",
+  "caramelised": "caramelized",
+  "caramelises": "caramelizes",
+  "caramelising": "caramelizing",
+  "carbonise": "carbonize",
+  "carbonised": "carbonized",
+  "carbonises": "carbonizes",
+  "carbonising": "carbonizing",
+  "carolled": "caroled",
+  "carolling": "caroling",
+  "catalogue": "catalog",
+  "catalogued": "cataloged",
+  "catalogues": "catalogs",
+  "cataloguing": "cataloging",
+  "catalyse": "catalyze",
+  "catalysed": "catalyzed",
+  "catalyses": "catalyzes",
+  "catalysing": "catalyzing",
+  "categorise": "categorize",
+  "categorised": "categorized",
+  "categorises": "categorizes",
+  "categorising": "categorizing",
+  "cauterise": "cauterize",
+  "cauterised": "cauterized",
+  "cauterises": "cauterizes",
+  "cauterising": "cauterizing",
+  "cavilled": "caviled",
+  "cavilling": "caviling",
+  "centigramme": "centigram",
+  "centigrammes": "centigrams",
+  "centilitre": "centiliter",
+  "centilitres": "centiliters",
+  "centimetre": "centimeter",
+  "centimetres": "centimeters",
+  "centralise": "centralize",
+  "centralised": "centralized",
+  "centralises": "centralizes",
+  "centralising": "centralizing",
+  "centre": "center",
+  "centred": "centered",
+  "centrefold": "centerfold",
+  "centrefolds": "centerfolds",
+  "centrepiece": "centerpiece",
+  "centrepieces": "centerpieces",
+  "centres": "centers",
+  "channelled": "channeled",
+  "channelling": "channeling",
+  "characterise": "characterize",
+  "characterised": "characterized",
+  "characterises": "characterizes",
+  "characterising": "characterizing",
+  "cheque": "check",
+  "chequebook": "checkbook",
+  "chequebooks": "checkbooks",
+  "chequered": "checkered",
+  "cheques": "checks",
+  "chilli": "chili",
+  "chimaera": "chimera",
+  "chimaeras": "chimeras",
+  "chiselled": "chiseled",
+  "chiselling": "chiseling",
+  "circularise": "circularize",
+  "circularised": "circularized",
+  "circularises": "circularizes",
+  "circularising": "circularizing",
+  "civilise": "civilize",
+  "civilised": "civilized",
+  "civilises": "civilizes",
+  "civilising": "civilizing",
+  "clamour": "clamor",
+  "clamoured": "clamored",
+  "clamouring": "clamoring",
+  "clamours": "clamors",
+  "clangour": "clangor",
+  "clarinettist": "clarinetist",
+  "clarinettists": "clarinetists",
+  "collectivise": "collectivize",
+  "collectivised": "collectivized",
+  "collectivises": "collectivizes",
+  "collectivising": "collectivizing",
+  "colonisation": "colonization",
+  "colonise": "colonize",
+  "colonised": "colonized",
+  "coloniser": "colonizer",
+  "colonisers": "colonizers",
+  "colonises": "colonizes",
+  "colonising": "colonizing",
+  "colour": "color",
+  "colourant": "colorant",
+  "colourants": "colorants",
+  "coloured": "colored",
+  "coloureds": "coloreds",
+  "colourful": "colorful",
+  "colourfully": "colorfully",
+  "colouring": "coloring",
+  "colourize": "colorize",
+  "colourized": "colorized",
+  "colourizes": "colorizes",
+  "colourizing": "colorizing",
+  "colourless": "colorless",
+  "colours": "colors",
+  "commercialise": "commercialize",
+  "commercialised": "commercialized",
+  "commercialises": "commercializes",
+  "commercialising": "commercializing",
+  "compartmentalise": "compartmentalize",
+  "compartmentalised": "compartmentalized",
+  "compartmentalises": "compartmentalizes",
+  "compartmentalising": "compartmentalizing",
+  "computerise": "computerize",
+  "computerised": "computerized",
+  "computerises": "computerizes",
+  "computerising": "computerizing",
+  "conceptualise": "conceptualize",
+  "conceptualised": "conceptualized",
+  "conceptualises": "conceptualizes",
+  "conceptualising": "conceptualizing",
+  "connexion": "connection",
+  "connexions": "connections",
+  "contextualise": "contextualize",
+  "contextualised": "contextualized",
+  "contextualises": "contextualizes",
+  "contextualising": "contextualizing",
+  "cosier": "cozier",
+  "cosies": "cozies",
+  "cosiest": "coziest",
+  "cosily": "cozily",
+  "cosiness": "coziness",
+  "cosy": "cozy",
+  "councillor": "councilor",
+  "councillors": "councilors",
+  "counselled": "counseled",
+  "counselling": "counseling",
+  "counsellor": "counselor",
+  "counsellors": "counselors",
+  "crenelated": "crenellated",
+  "criminalise": "criminalize",
+  "criminalised": "criminalized",
+  "criminalises": "criminalizes",
+  "criminalising": "criminalizing",
+  "criticise": "criticize",
+  "criticised": "criticized",
+  "criticises": "criticizes",
+  "criticising": "criticizing",
+  "crueller": "crueler",
+  "cruellest": "cruelest",
+  "crystallisation": "crystallization",
+  "crystallise": "crystallize",
+  "crystallised": "crystallized",
+  "crystallises": "crystallizes",
+  "crystallising": "crystallizing",
+  "cudgelled": "cudgeled",
+  "cudgelling": "cudgeling",
+  "customise": "customize",
+  "customised": "customized",
+  "customises": "customizes",
+  "customising": "customizing",
+  "cypher": "cipher",
+  "cyphers": "ciphers",
+  "decentralisation": "decentralization",
+  "decentralise": "decentralize",
+  "decentralised": "decentralized",
+  "decentralises": "decentralizes",
+  "decentralising": "decentralizing",
+  "decriminalisation": "decriminalization",
+  "decriminalise": "decriminalize",
+  "decriminalised": "decriminalized",
+  "decriminalises": "decriminalizes",
+  "decriminalising": "decriminalizing",
+  "defence": "defense",
+  "defenceless": "defenseless",
+  "defences": "defenses",
+  "dehumanisation": "dehumanization",
+  "dehumanise": "dehumanize",
+  "dehumanised": "dehumanized",
+  "dehumanises": "dehumanizes",
+  "dehumanising": "dehumanizing",
+  "demeanour": "demeanor",
+  "demilitarisation": "demilitarization",
+  "demilitarise": "demilitarize",
+  "demilitarised": "demilitarized",
+  "demilitarises": "demilitarizes",
+  "demilitarising": "demilitarizing",
+  "demobilisation": "demobilization",
+  "demobilise": "demobilize",
+  "demobilised": "demobilized",
+  "demobilises": "demobilizes",
+  "demobilising": "demobilizing",
+  "democratisation": "democratization",
+  "democratise": "democratize",
+  "democratised": "democratized",
+  "democratises": "democratizes",
+  "democratising": "democratizing",
+  "demonise": "demonize",
+  "demonised": "demonized",
+  "demonises": "demonizes",
+  "demonising": "demonizing",
+  "demoralisation": "demoralization",
+  "demoralise": "demoralize",
+  "demoralised": "demoralized",
+  "demoralises": "demoralizes",
+  "demoralising": "demoralizing",
+  "denationalisation": "denationalization",
+  "denationalise": "denationalize",
+  "denationalised": "denationalized",
+  "denationalises": "denationalizes",
+  "denationalising": "denationalizing",
+  "deodorise": "deodorize",
+  "deodorised": "deodorized",
+  "deodorises": "deodorizes",
+  "deodorising": "deodorizing",
+  "depersonalise": "depersonalize",
+  "depersonalised": "depersonalized",
+  "depersonalises": "depersonalizes",
+  "depersonalising": "depersonalizing",
+  "deputise": "deputize",
+  "deputised": "deputized",
+  "deputises": "deputizes",
+  "deputising": "deputizing",
+  "desensitisation": "desensitization",
+  "desensitise": "desensitize",
+  "desensitised": "desensitized",
+  "desensitises": "desensitizes",
+  "desensitising": "desensitizing",
+  "destabilisation": "destabilization",
+  "destabilise": "destabilize",
+  "destabilised": "destabilized",
+  "destabilises": "destabilizes",
+  "destabilising": "destabilizing",
+  "dialled": "dialed",
+  "dialling": "dialing",
+  "dialogue": "dialog",
+  "dialogues": "dialogs",
+  "diarrhoea": "diarrhea",
+  "digitise": "digitize",
+  "digitised": "digitized",
+  "digitises": "digitizes",
+  "digitising": "digitizing",
+  "disc": "disk",
+  "discolour": "discolor",
+  "discoloured": "discolored",
+  "discolouring": "discoloring",
+  "discolours": "discolors",
+  "discs": "disks",
+  "disembowelled": "disemboweled",
+  "disembowelling": "disemboweling",
+  "disfavour": "disfavor",
+  "dishevelled": "disheveled",
+  "dishonour": "dishonor",
+  "dishonourable": "dishonorable",
+  "dishonourably": "dishonorably",
+  "dishonoured": "dishonored",
+  "dishonouring": "dishonoring",
+  "dishonours": "dishonors",
+  "disorganisation": "disorganization",
+  "disorganised": "disorganized",
+  "distil": "distill",
+  "distils": "distills",
+  "dramatisation": "dramatization",
+  "dramatisations": "dramatizations",
+  "dramatise": "dramatize",
+  "dramatised": "dramatized",
+  "dramatises": "dramatizes",
+  "dramatising": "dramatizing",
+  "draught": "draft",
+  "draughtboard": "draftboard",
+  "draughtboards": "draftboards",
+  "draughtier": "draftier",
+  "draughtiest": "draftiest",
+  "draughts": "drafts",
+  "draughtsman": "draftsman",
+  "draughtsmanship": "draftsmanship",
+  "draughtsmen": "draftsmen",
+  "draughtswoman": "draftswoman",
+  "draughtswomen": "draftswomen",
+  "draughty": "drafty",
+  "drivelled": "driveled",
+  "drivelling": "driveling",
+  "duelled": "dueled",
+  "duelling": "dueling",
+  "economise": "economize",
+  "economised": "economized",
+  "economises": "economizes",
+  "economising": "economizing",
+  "editorialise": "editorialize",
+  "editorialised": "editorialized",
+  "editorialises": "editorializes",
+  "editorialising": "editorializing",
+  "edoema": "edema",
+  "empathise": "empathize",
+  "empathised": "empathized",
+  "empathises": "empathizes",
+  "empathising": "empathizing",
+  "emphasise": "emphasize",
+  "emphasised": "emphasized",
+  "emphasises": "emphasizes",
+  "emphasising": "emphasizing",
+  "enamelled": "enameled",
+  "enamelling": "enameling",
+  "enamoured": "enamored",
+  "encyclopaedia": "encyclopedia",
+  "encyclopaedias": "encyclopedias",
+  "encyclopaedic": "encyclopedic",
+  "endeavour": "endeavor",
+  "endeavoured": "endeavored",
+  "endeavouring": "endeavoring",
+  "endeavours": "endeavors",
+  "energise": "energize",
+  "energised": "energized",
+  "energises": "energizes",
+  "energising": "energizing",
+  "enrol": "enroll",
+  "enrols": "enrolls",
+  "enthral": "enthrall",
+  "enthrals": "enthralls",
+  "epaulette": "epaulet",
+  "epaulettes": "epaulets",
+  "epicentre": "epicenter",
+  "epicentres": "epicenters",
+  "epilogue": "epilog",
+  "epilogues": "epilogs",
+  "epitomise": "epitomize",
+  "epitomised": "epitomized",
+  "epitomises": "epitomizes",
+  "epitomising": "epitomizing",
+  "equalisation": "equalization",
+  "equalise": "equalize",
+  "equalised": "equalized",
+  "equaliser": "equalizer",
+  "equalisers": "equalizers",
+  "equalises": "equalizes",
+  "equalising": "equalizing",
+  "eulogise": "eulogize",
+  "eulogised": "eulogized",
+  "eulogises": "eulogizes",
+  "eulogising": "eulogizing",
+  "evangelise": "evangelize",
+  "evangelised": "evangelized",
+  "evangelises": "evangelizes",
+  "evangelising": "evangelizing",
+  "exorcise": "exorcize",
+  "exorcised": "exorcized",
+  "exorcises": "exorcizes",
+  "exorcising": "exorcizing",
+  "extemporisation": "extemporization",
+  "extemporise": "extemporize",
+  "extemporised": "extemporized",
+  "extemporises": "extemporizes",
+  "extemporising": "extemporizing",
+  "externalisation": "externalization",
+  "externalisations": "externalizations",
+  "externalise": "externalize",
+  "externalised": "externalized",
+  "externalises": "externalizes",
+  "externalising": "externalizing",
+  "factorise": "factorize",
+  "factorised": "factorized",
+  "factorises": "factorizes",
+  "factorising": "factorizing",
+  "faecal": "fecal",
+  "faeces": "feces",
+  "familiarisation": "familiarization",
+  "familiarise": "familiarize",
+  "familiarised": "familiarized",
+  "familiarises": "familiarizes",
+  "familiarising": "familiarizing",
+  "fantasise": "fantasize",
+  "fantasised": "fantasized",
+  "fantasises": "fantasizes",
+  "fantasising": "fantasizing",
+  "favour": "favor",
+  "favourable": "favorable",
+  "favourably": "favorably",
+  "favoured": "favored",
+  "favouring": "favoring",
+  "favourite": "favorite",
+  "favourites": "favorites",
+  "favouritism": "favoritism",
+  "favours": "favors",
+  "feminise": "feminize",
+  "feminised": "feminized",
+  "feminises": "feminizes",
+  "feminising": "feminizing",
+  "fertilisation": "fertilization",
+  "fertilise": "fertilize",
+  "fertilised": "fertilized",
+  "fertiliser": "fertilizer",
+  "fertilisers": "fertilizers",
+  "fertilises": "fertilizes",
+  "fertilising": "fertilizing",
+  "fervour": "fervor",
+  "fibre": "fiber",
+  "fibreglass": "fiberglass",
+  "fibres": "fibers",
+  "fictionalisation": "fictionalization",
+  "fictionalisations": "fictionalizations",
+  "fictionalise": "fictionalize",
+  "fictionalised": "fictionalized",
+  "fictionalises": "fictionalizes",
+  "fictionalising": "fictionalizing",
+  "fillet": "filet",
+  "filleted": "fileted",
+  "filleting": "fileting",
+  "fillets": "filets",
+  "finalisation": "finalization",
+  "finalise": "finalize",
+  "finalised": "finalized",
+  "finalises": "finalizes",
+  "finalising": "finalizing",
+  "flautist": "flutist",
+  "flautists": "flutists",
+  "flavour": "flavor",
+  "flavoured": "flavored",
+  "flavouring": "flavoring",
+  "flavourings": "flavorings",
+  "flavourless": "flavorless",
+  "flavours": "flavors",
+  "flavoursome": "flavorsome",
+  "flyer / flier": "flier / flyer",
+  "foetal": "fetal",
+  "foetid": "fetid",
+  "foetus": "fetus",
+  "foetuses": "fetuses",
+  "formalisation": "formalization",
+  "formalise": "formalize",
+  "formalised": "formalized",
+  "formalises": "formalizes",
+  "formalising": "formalizing",
+  "fossilisation": "fossilization",
+  "fossilise": "fossilize",
+  "fossilised": "fossilized",
+  "fossilises": "fossilizes",
+  "fossilising": "fossilizing",
+  "fraternisation": "fraternization",
+  "fraternise": "fraternize",
+  "fraternised": "fraternized",
+  "fraternises": "fraternizes",
+  "fraternising": "fraternizing",
+  "fulfil": "fulfill",
+  "fulfilment": "fulfillment",
+  "fulfils": "fulfills",
+  "funnelled": "funneled",
+  "funnelling": "funneling",
+  "gage": "gauge",
+  "gaged": "gauged",
+  "gages": "gauges",
+  "gaging": "gauging",
+  "galvanise": "galvanize",
+  "galvanised": "galvanized",
+  "galvanises": "galvanizes",
+  "galvanising": "galvanizing",
+  "gambolled": "gamboled",
+  "gambolling": "gamboling",
+  "gaol": "jail",
+  "gaolbird": "jailbird",
+  "gaolbirds": "jailbirds",
+  "gaolbreak": "jailbreak",
+  "gaolbreaks": "jailbreaks",
+  "gaoled": "jailed",
+  "gaoler": "jailer",
+  "gaolers": "jailers",
+  "gaoling": "jailing",
+  "gaols": "jails",
+  "gasses": "gases",
+  "generalisation": "generalization",
+  "generalisations": "generalizations",
+  "generalise": "generalize",
+  "generalised": "generalized",
+  "generalises": "generalizes",
+  "generalising": "generalizing",
+  "ghettoise": "ghettoize",
+  "ghettoised": "ghettoized",
+  "ghettoises": "ghettoizes",
+  "ghettoising": "ghettoizing",
+  "gipsies": "gypsies",
+  "glamor": "glamour",
+  "glamorise": "glamorize",
+  "glamorised": "glamorized",
+  "glamorises": "glamorizes",
+  "glamorising": "glamorizing",
+  "globalisation": "globalization",
+  "globalise": "globalize",
+  "globalised": "globalized",
+  "globalises": "globalizes",
+  "globalising": "globalizing",
+  "glueing": "gluing",
+  "goitre": "goiter",
+  "goitres": "goiters",
+  "gonorrhoea": "gonorrhea",
+  "gramme": "gram",
+  "grammes": "grams",
+  "gravelled": "graveled",
+  "grey": "gray",
+  "greyed": "grayed",
+  "greying": "graying",
+  "greyish": "grayish",
+  "greyness": "grayness",
+  "greys": "grays",
+  "grovelled": "groveled",
+  "grovelling": "groveling",
+  "groyne": "groin",
+  "groynes": "groins",
+  "gruelling": "grueling",
+  "gruellingly": "gruelingly",
+  "gryphon": "griffin",
+  "gryphons": "griffins",
+  "gynaecological": "gynecological",
+  "gynaecologist": "gynecologist",
+  "gynaecologists": "gynecologists",
+  "gynaecology": "gynecology",
+  "haematological": "hematological",
+  "haematologist": "hematologist",
+  "haematologists": "hematologists",
+  "haematology": "hematology",
+  "haemoglobin": "hemoglobin",
+  "haemophilia": "hemophilia",
+  "haemophiliac": "hemophiliac",
+  "haemophiliacs": "hemophiliacs",
+  "haemorrhage": "hemorrhage",
+  "haemorrhaged": "hemorrhaged",
+  "haemorrhages": "hemorrhages",
+  "haemorrhaging": "hemorrhaging",
+  "haemorrhoids": "hemorrhoids",
+  "harbour": "harbor",
+  "harboured": "harbored",
+  "harbouring": "harboring",
+  "harbours": "harbors",
+  "harmonisation": "harmonization",
+  "harmonise": "harmonize",
+  "harmonised": "harmonized",
+  "harmonises": "harmonizes",
+  "harmonising": "harmonizing",
+  "homoeopath": "homeopath",
+  "homoeopathic": "homeopathic",
+  "homoeopaths": "homeopaths",
+  "homoeopathy": "homeopathy",
+  "homogenise": "homogenize",
+  "homogenised": "homogenized",
+  "homogenises": "homogenizes",
+  "homogenising": "homogenizing",
+  "honour": "honor",
+  "honourable": "honorable",
+  "honourably": "honorably",
+  "honoured": "honored",
+  "honouring": "honoring",
+  "honours": "honors",
+  "hospitalisation": "hospitalization",
+  "hospitalise": "hospitalize",
+  "hospitalised": "hospitalized",
+  "hospitalises": "hospitalizes",
+  "hospitalising": "hospitalizing",
+  "humanise": "humanize",
+  "humanised": "humanized",
+  "humanises": "humanizes",
+  "humanising": "humanizing",
+  "humour": "humor",
+  "humoured": "humored",
+  "humouring": "humoring",
+  "humourless": "humorless",
+  "humours": "humors",
+  "hybridise": "hybridize",
+  "hybridised": "hybridized",
+  "hybridises": "hybridizes",
+  "hybridising": "hybridizing",
+  "hypnotise": "hypnotize",
+  "hypnotised": "hypnotized",
+  "hypnotises": "hypnotizes",
+  "hypnotising": "hypnotizing",
+  "hypothesise": "hypothesize",
+  "hypothesised": "hypothesized",
+  "hypothesises": "hypothesizes",
+  "hypothesising": "hypothesizing",
+  "idealisation": "idealization",
+  "idealise": "idealize",
+  "idealised": "idealized",
+  "idealises": "idealizes",
+  "idealising": "idealizing",
+  "idolise": "idolize",
+  "idolised": "idolized",
+  "idolises": "idolizes",
+  "idolising": "idolizing",
+  "immobilisation": "immobilization",
+  "immobilise": "immobilize",
+  "immobilised": "immobilized",
+  "immobiliser": "immobilizer",
+  "immobilisers": "immobilizers",
+  "immobilises": "immobilizes",
+  "immobilising": "immobilizing",
+  "immortalise": "immortalize",
+  "immortalised": "immortalized",
+  "immortalises": "immortalizes",
+  "immortalising": "immortalizing",
+  "immunisation": "immunization",
+  "immunise": "immunize",
+  "immunised": "immunized",
+  "immunises": "immunizes",
+  "immunising": "immunizing",
+  "impanelled": "impaneled",
+  "impanelling": "impaneling",
+  "imperilled": "imperiled",
+  "imperilling": "imperiling",
+  "individualise": "individualize",
+  "individualised": "individualized",
+  "individualises": "individualizes",
+  "individualising": "individualizing",
+  "industrialise": "industrialize",
+  "industrialised": "industrialized",
+  "industrialises": "industrializes",
+  "industrialising": "industrializing",
+  "inflexion": "inflection",
+  "inflexions": "inflections",
+  "initialise": "initialize",
+  "initialised": "initialized",
+  "initialises": "initializes",
+  "initialising": "initializing",
+  "initialled": "initialed",
+  "initialling": "initialing",
+  "instal": "install",
+  "instalment": "installment",
+  "instalments": "installments",
+  "instals": "installs",
+  "instil": "instill",
+  "instils": "instills",
+  "institutionalisation": "institutionalization",
+  "institutionalise": "institutionalize",
+  "institutionalised": "institutionalized",
+  "institutionalises": "institutionalizes",
+  "institutionalising": "institutionalizing",
+  "intellectualise": "intellectualize",
+  "intellectualised": "intellectualized",
+  "intellectualises": "intellectualizes",
+  "intellectualising": "intellectualizing",
+  "internalisation": "internalization",
+  "internalise": "internalize",
+  "internalised": "internalized",
+  "internalises": "internalizes",
+  "internalising": "internalizing",
+  "internationalisation": "internationalization",
+  "internationalise": "internationalize",
+  "internationalised": "internationalized",
+  "internationalises": "internationalizes",
+  "internationalising": "internationalizing",
+  "ionisation": "ionization",
+  "ionise": "ionize",
+  "ionised": "ionized",
+  "ioniser": "ionizer",
+  "ionisers": "ionizers",
+  "ionises": "ionizes",
+  "ionising": "ionizing",
+  "italicise": "italicize",
+  "italicised": "italicized",
+  "italicises": "italicizes",
+  "italicising": "italicizing",
+  "itemise": "itemize",
+  "itemised": "itemized",
+  "itemises": "itemizes",
+  "itemising": "itemizing",
+  "jeopardise": "jeopardize",
+  "jeopardised": "jeopardized",
+  "jeopardises": "jeopardizes",
+  "jeopardising": "jeopardizing",
+  "jewelled": "jeweled",
+  "jeweller": "jeweler",
+  "jewellers": "jewelers",
+  "jewellery": "jewelry",
+  "judgement": "judgment",
+  "kilogramme": "kilogram",
+  "kilogrammes": "kilograms",
+  "kilometre": "kilometer",
+  "kilometres": "kilometers",
+  "labelled": "labeled",
+  "labelling": "labeling",
+  "labour": "labor",
+  "laboured": "labored",
+  "labourer": "laborer",
+  "labourers": "laborers",
+  "labouring": "laboring",
+  "labours": "labors",
+  "lacklustre": "lackluster",
+  "legalisation": "legalization",
+  "legalise": "legalize",
+  "legalised": "legalized",
+  "legalises": "legalizes",
+  "legalising": "legalizing",
+  "legitimise": "legitimize",
+  "legitimised": "legitimized",
+  "legitimises": "legitimizes",
+  "legitimising": "legitimizing",
+  "leukaemia": "leukemia",
+  "levelled": "leveled",
+  "leveller": "leveler",
+  "levellers": "levelers",
+  "levelling": "leveling",
+  "libelled": "libeled",
+  "libelling": "libeling",
+  "libellous": "libelous",
+  "liberalisation": "liberalization",
+  "liberalise": "liberalize",
+  "liberalised": "liberalized",
+  "liberalises": "liberalizes",
+  "liberalising": "liberalizing",
+  "licence": "license",
+  "licenced": "licensed",
+  "licences": "licenses",
+  "licencing": "licensing",
+  "likeable": "likable",
+  "lionisation": "lionization",
+  "lionise": "lionize",
+  "lionised": "lionized",
+  "lionises": "lionizes",
+  "lionising": "lionizing",
+  "liquidise": "liquidize",
+  "liquidised": "liquidized",
+  "liquidiser": "liquidizer",
+  "liquidisers": "liquidizers",
+  "liquidises": "liquidizes",
+  "liquidising": "liquidizing",
+  "litre": "liter",
+  "litres": "liters",
+  "localise": "localize",
+  "localised": "localized",
+  "localises": "localizes",
+  "localising": "localizing",
+  "louvre": "louver",
+  "louvred": "louvered",
+  "louvres": "louvers",
+  "lustre": "luster",
+  "magnetise": "magnetize",
+  "magnetised": "magnetized",
+  "magnetises": "magnetizes",
+  "magnetising": "magnetizing",
+  "manoeuvrability": "maneuverability",
+  "manoeuvrable": "maneuverable",
+  "manoeuvre": "maneuver",
+  "manoeuvred": "maneuvered",
+  "manoeuvres": "maneuvers",
+  "manoeuvring": "maneuvering",
+  "manoeuvrings": "maneuverings",
+  "marginalisation": "marginalization",
+  "marginalise": "marginalize",
+  "marginalised": "marginalized",
+  "marginalises": "marginalizes",
+  "marginalising": "marginalizing",
+  "marshalled": "marshaled",
+  "marshalling": "marshaling",
+  "marvelled": "marveled",
+  "marvelling": "marveling",
+  "marvellous": "marvelous",
+  "marvellously": "marvelously",
+  "materialisation": "materialization",
+  "materialise": "materialize",
+  "materialised": "materialized",
+  "materialises": "materializes",
+  "materialising": "materializing",
+  "maximisation": "maximization",
+  "maximise": "maximize",
+  "maximised": "maximized",
+  "maximises": "maximizes",
+  "maximising": "maximizing",
+  "meagre": "meager",
+  "mechanisation": "mechanization",
+  "mechanise": "mechanize",
+  "mechanised": "mechanized",
+  "mechanises": "mechanizes",
+  "mechanising": "mechanizing",
+  "mediaeval": "medieval",
+  "memorialise": "memorialize",
+  "memorialised": "memorialized",
+  "memorialises": "memorializes",
+  "memorialising": "memorializing",
+  "memorise": "memorize",
+  "memorised": "memorized",
+  "memorises": "memorizes",
+  "memorising": "memorizing",
+  "mesmerise": "mesmerize",
+  "mesmerised": "mesmerized",
+  "mesmerises": "mesmerizes",
+  "mesmerising": "mesmerizing",
+  "metabolise": "metabolize",
+  "metabolised": "metabolized",
+  "metabolises": "metabolizes",
+  "metabolising": "metabolizing",
+  "metre": "meter",
+  "metres": "meters",
+  "mhm": "hmm",
+  "micrometre": "micrometer",
+  "micrometres": "micrometers",
+  "militarise": "militarize",
+  "militarised": "militarized",
+  "militarises": "militarizes",
+  "militarising": "militarizing",
+  "milligramme": "milligram",
+  "milligrammes": "milligrams",
+  "millilitre": "milliliter",
+  "millilitres": "milliliters",
+  "millimetre": "millimeter",
+  "millimetres": "millimeters",
+  "miniaturisation": "miniaturization",
+  "miniaturise": "miniaturize",
+  "miniaturised": "miniaturized",
+  "miniaturises": "miniaturizes",
+  "miniaturising": "miniaturizing",
+  "minibusses": "minibuses",
+  "minimise": "minimize",
+  "minimised": "minimized",
+  "minimises": "minimizes",
+  "minimising": "minimizing",
+  "misbehaviour": "misbehavior",
+  "misdemeanour": "misdemeanor",
+  "misdemeanours": "misdemeanors",
+  "misspelt": "misspelled",
+  "mitre": "miter",
+  "mitres": "miters",
+  "mm": "hmm",
+  "mmm": "hmm",
+  "mobilisation": "mobilization",
+  "mobilise": "mobilize",
+  "mobilised": "mobilized",
+  "mobilises": "mobilizes",
+  "mobilising": "mobilizing",
+  "modelled": "modeled",
+  "modeller": "modeler",
+  "modellers": "modelers",
+  "modelling": "modeling",
+  "modernise": "modernize",
+  "modernised": "modernized",
+  "modernises": "modernizes",
+  "modernising": "modernizing",
+  "moisturise": "moisturize",
+  "moisturised": "moisturized",
+  "moisturiser": "moisturizer",
+  "moisturisers": "moisturizers",
+  "moisturises": "moisturizes",
+  "moisturising": "moisturizing",
+  "monologue": "monolog",
+  "monologues": "monologs",
+  "monopolisation": "monopolization",
+  "monopolise": "monopolize",
+  "monopolised": "monopolized",
+  "monopolises": "monopolizes",
+  "monopolising": "monopolizing",
+  "moralise": "moralize",
+  "moralised": "moralized",
+  "moralises": "moralizes",
+  "moralising": "moralizing",
+  "motorised": "motorized",
+  "mould": "mold",
+  "moulded": "molded",
+  "moulder": "molder",
+  "mouldered": "moldered",
+  "mouldering": "moldering",
+  "moulders": "molders",
+  "mouldier": "moldier",
+  "mouldiest": "moldiest",
+  "moulding": "molding",
+  "mouldings": "moldings",
+  "moulds": "molds",
+  "mouldy": "moldy",
+  "moult": "molt",
+  "moulted": "molted",
+  "moulting": "molting",
+  "moults": "molts",
+  "moustache": "mustache",
+  "moustached": "mustached",
+  "moustaches": "mustaches",
+  "moustachioed": "mustachioed",
+  "multicoloured": "multicolored",
+  "nationalisation": "nationalization",
+  "nationalisations": "nationalizations",
+  "nationalise": "nationalize",
+  "nationalised": "nationalized",
+  "nationalises": "nationalizes",
+  "nationalising": "nationalizing",
+  "naturalisation": "naturalization",
+  "naturalise": "naturalize",
+  "naturalised": "naturalized",
+  "naturalises": "naturalizes",
+  "naturalising": "naturalizing",
+  "neighbour": "neighbor",
+  "neighbourhood": "neighborhood",
+  "neighbourhoods": "neighborhoods",
+  "neighbouring": "neighboring",
+  "neighbourliness": "neighborliness",
+  "neighbourly": "neighborly",
+  "neighbours": "neighbors",
+  "neutralisation": "neutralization",
+  "neutralise": "neutralize",
+  "neutralised": "neutralized",
+  "neutralises": "neutralizes",
+  "neutralising": "neutralizing",
+  "normalisation": "normalization",
+  "normalise": "normalize",
+  "normalised": "normalized",
+  "normalises": "normalizes",
+  "normalising": "normalizing",
+  "odour": "odor",
+  "odourless": "odorless",
+  "odours": "odors",
+  "oesophagus": "esophagus",
+  "oesophaguses": "esophaguses",
+  "oestrogen": "estrogen",
+  "offence": "offense",
+  "offences": "offenses",
+  "omelette": "omelet",
+  "omelettes": "omelets",
+  "optimise": "optimize",
+  "optimised": "optimized",
+  "optimises": "optimizes",
+  "optimising": "optimizing",
+  "organisation": "organization",
+  "organisational": "organizational",
+  "organisations": "organizations",
+  "organise": "organize",
+  "organised": "organized",
+  "organiser": "organizer",
+  "organisers": "organizers",
+  "organises": "organizes",
+  "organising": "organizing",
+  "orthopaedic": "orthopedic",
+  "orthopaedics": "orthopedics",
+  "ostracise": "ostracize",
+  "ostracised": "ostracized",
+  "ostracises": "ostracizes",
+  "ostracising": "ostracizing",
+  "outmanoeuvre": "outmaneuver",
+  "outmanoeuvred": "outmaneuvered",
+  "outmanoeuvres": "outmaneuvers",
+  "outmanoeuvring": "outmaneuvering",
+  "overemphasise": "overemphasize",
+  "overemphasised": "overemphasized",
+  "overemphasises": "overemphasizes",
+  "overemphasising": "overemphasizing",
+  "oxidisation": "oxidization",
+  "oxidise": "oxidize",
+  "oxidised": "oxidized",
+  "oxidises": "oxidizes",
+  "oxidising": "oxidizing",
+  "paederast": "pederast",
+  "paederasts": "pederasts",
+  "paediatric": "pediatric",
+  "paediatrician": "pediatrician",
+  "paediatricians": "pediatricians",
+  "paediatrics": "pediatrics",
+  "paedophile": "pedophile",
+  "paedophiles": "pedophiles",
+  "paedophilia": "pedophilia",
+  "palaeolithic": "paleolithic",
+  "palaeontologist": "paleontologist",
+  "palaeontologists": "paleontologists",
+  "palaeontology": "paleontology",
+  "panelled": "paneled",
+  "panelling": "paneling",
+  "panellist": "panelist",
+  "panellists": "panelists",
+  "paralyse": "paralyze",
+  "paralysed": "paralyzed",
+  "paralyses": "paralyzes",
+  "paralysing": "paralyzing",
+  "parcelled": "parceled",
+  "parcelling": "parceling",
+  "parlour": "parlor",
+  "parlours": "parlors",
+  "particularise": "particularize",
+  "particularised": "particularized",
+  "particularises": "particularizes",
+  "particularising": "particularizing",
+  "passivisation": "passivization",
+  "passivise": "passivize",
+  "passivised": "passivized",
+  "passivises": "passivizes",
+  "passivising": "passivizing",
+  "pasteurisation": "pasteurization",
+  "pasteurise": "pasteurize",
+  "pasteurised": "pasteurized",
+  "pasteurises": "pasteurizes",
+  "pasteurising": "pasteurizing",
+  "patronise": "patronize",
+  "patronised": "patronized",
+  "patronises": "patronizes",
+  "patronising": "patronizing",
+  "patronisingly": "patronizingly",
+  "pedalled": "pedaled",
+  "pedalling": "pedaling",
+  "pedestrianisation": "pedestrianization",
+  "pedestrianise": "pedestrianize",
+  "pedestrianised": "pedestrianized",
+  "pedestrianises": "pedestrianizes",
+  "pedestrianising": "pedestrianizing",
+  "penalise": "penalize",
+  "penalised": "penalized",
+  "penalises": "penalizes",
+  "penalising": "penalizing",
+  "pencilled": "penciled",
+  "pencilling": "penciling",
+  "personalise": "personalize",
+  "personalised": "personalized",
+  "personalises": "personalizes",
+  "personalising": "personalizing",
+  "pharmacopoeia": "pharmacopeia",
+  "pharmacopoeias": "pharmacopeias",
+  "philosophise": "philosophize",
+  "philosophised": "philosophized",
+  "philosophises": "philosophizes",
+  "philosophising": "philosophizing",
+  "philtre": "filter",
+  "philtres": "filters",
+  "phoney": "phony",
+  "plagiarise": "plagiarize",
+  "plagiarised": "plagiarized",
+  "plagiarises": "plagiarizes",
+  "plagiarising": "plagiarizing",
+  "plough": "plow",
+  "ploughed": "plowed",
+  "ploughing": "plowing",
+  "ploughman": "plowman",
+  "ploughmen": "plowmen",
+  "ploughs": "plows",
+  "ploughshare": "plowshare",
+  "ploughshares": "plowshares",
+  "polarisation": "polarization",
+  "polarise": "polarize",
+  "polarised": "polarized",
+  "polarises": "polarizes",
+  "polarising": "polarizing",
+  "politicisation": "politicization",
+  "politicise": "politicize",
+  "politicised": "politicized",
+  "politicises": "politicizes",
+  "politicising": "politicizing",
+  "popularisation": "popularization",
+  "popularise": "popularize",
+  "popularised": "popularized",
+  "popularises": "popularizes",
+  "popularising": "popularizing",
+  "pouffe": "pouf",
+  "pouffes": "poufs",
+  "practise": "practice",
+  "practised": "practiced",
+  "practises": "practices",
+  "practising": "practicing",
+  "praesidium": "presidium",
+  "praesidiums": "presidiums",
+  "pressurisation": "pressurization",
+  "pressurise": "pressurize",
+  "pressurised": "pressurized",
+  "pressurises": "pressurizes",
+  "pressurising": "pressurizing",
+  "pretence": "pretense",
+  "pretences": "pretenses",
+  "primaeval": "primeval",
+  "prioritisation": "prioritization",
+  "prioritise": "prioritize",
+  "prioritised": "prioritized",
+  "prioritises": "prioritizes",
+  "prioritising": "prioritizing",
+  "privatisation": "privatization",
+  "privatisations": "privatizations",
+  "privatise": "privatize",
+  "privatised": "privatized",
+  "privatises": "privatizes",
+  "privatising": "privatizing",
+  "professionalisation": "professionalization",
+  "professionalise": "professionalize",
+  "professionalised": "professionalized",
+  "professionalises": "professionalizes",
+  "professionalising": "professionalizing",
+  "programme": "program",
+  "programmes": "programs",
+  "prologue": "prolog",
+  "prologues": "prologs",
+  "propagandise": "propagandize",
+  "propagandised": "propagandized",
+  "propagandises": "propagandizes",
+  "propagandising": "propagandizing",
+  "proselytise": "proselytize",
+  "proselytised": "proselytized",
+  "proselytiser": "proselytizer",
+  "proselytisers": "proselytizers",
+  "proselytises": "proselytizes",
+  "proselytising": "proselytizing",
+  "psychoanalyse": "psychoanalyze",
+  "psychoanalysed": "psychoanalyzed",
+  "psychoanalyses": "psychoanalyzes",
+  "psychoanalysing": "psychoanalyzing",
+  "publicise": "publicize",
+  "publicised": "publicized",
+  "publicises": "publicizes",
+  "publicising": "publicizing",
+  "pulverisation": "pulverization",
+  "pulverise": "pulverize",
+  "pulverised": "pulverized",
+  "pulverises": "pulverizes",
+  "pulverising": "pulverizing",
+  "pummelled": "pummel",
+  "pummelling": "pummeled",
+  "pyjama": "pajama",
+  "pyjamas": "pajamas",
+  "pzazz": "pizzazz",
+  "quarrelled": "quarreled",
+  "quarrelling": "quarreling",
+  "radicalise": "radicalize",
+  "radicalised": "radicalized",
+  "radicalises": "radicalizes",
+  "radicalising": "radicalizing",
+  "rancour": "rancor",
+  "randomise": "randomize",
+  "randomised": "randomized",
+  "randomises": "randomizes",
+  "randomising": "randomizing",
+  "rationalisation": "rationalization",
+  "rationalisations": "rationalizations",
+  "rationalise": "rationalize",
+  "rationalised": "rationalized",
+  "rationalises": "rationalizes",
+  "rationalising": "rationalizing",
+  "ravelled": "raveled",
+  "ravelling": "raveling",
+  "realisable": "realizable",
+  "realisation": "realization",
+  "realisations": "realizations",
+  "realise": "realize",
+  "realised": "realized",
+  "realises": "realizes",
+  "realising": "realizing",
+  "recognisable": "recognizable",
+  "recognisably": "recognizably",
+  "recognisance": "recognizance",
+  "recognise": "recognize",
+  "recognised": "recognized",
+  "recognises": "recognizes",
+  "recognising": "recognizing",
+  "reconnoitre": "reconnoiter",
+  "reconnoitred": "reconnoitered",
+  "reconnoitres": "reconnoiters",
+  "reconnoitring": "reconnoitering",
+  "refuelled": "refueled",
+  "refuelling": "refueling",
+  "regularisation": "regularization",
+  "regularise": "regularize",
+  "regularised": "regularized",
+  "regularises": "regularizes",
+  "regularising": "regularizing",
+  "remodelled": "remodeled",
+  "remodelling": "remodeling",
+  "remould": "remold",
+  "remoulded": "remolded",
+  "remoulding": "remolding",
+  "remoulds": "remolds",
+  "reorganisation": "reorganization",
+  "reorganisations": "reorganizations",
+  "reorganise": "reorganize",
+  "reorganised": "reorganized",
+  "reorganises": "reorganizes",
+  "reorganising": "reorganizing",
+  "revelled": "reveled",
+  "reveller": "reveler",
+  "revellers": "revelers",
+  "revelling": "reveling",
+  "revitalise": "revitalize",
+  "revitalised": "revitalized",
+  "revitalises": "revitalizes",
+  "revitalising": "revitalizing",
+  "revolutionise": "revolutionize",
+  "revolutionised": "revolutionized",
+  "revolutionises": "revolutionizes",
+  "revolutionising": "revolutionizing",
+  "rhapsodise": "rhapsodize",
+  "rhapsodised": "rhapsodized",
+  "rhapsodises": "rhapsodizes",
+  "rhapsodising": "rhapsodizing",
+  "rigour": "rigor",
+  "rigours": "rigors",
+  "ritualised": "ritualized",
+  "rivalled": "rivaled",
+  "rivalling": "rivaling",
+  "romanticise": "romanticize",
+  "romanticised": "romanticized",
+  "romanticises": "romanticizes",
+  "romanticising": "romanticizing",
+  "rumour": "rumor",
+  "rumoured": "rumored",
+  "rumours": "rumors",
+  "sabre": "saber",
+  "sabres": "sabers",
+  "saltpetre": "saltpeter",
+  "sanitise": "sanitize",
+  "sanitised": "sanitized",
+  "sanitises": "sanitizes",
+  "sanitising": "sanitizing",
+  "satirise": "satirize",
+  "satirised": "satirized",
+  "satirises": "satirizes",
+  "satirising": "satirizing",
+  "saviour": "savior",
+  "saviours": "saviors",
+  "savour": "savor",
+  "savoured": "savored",
+  "savouries": "savories",
+  "savouring": "savoring",
+  "savours": "savors",
+  "savoury": "savory",
+  "scandalise": "scandalize",
+  "scandalised": "scandalized",
+  "scandalises": "scandalizes",
+  "scandalising": "scandalizing",
+  "sceptic": "skeptic",
+  "sceptical": "skeptical",
+  "sceptically": "skeptically",
+  "scepticism": "skepticism",
+  "sceptics": "skeptics",
+  "sceptre": "scepter",
+  "sceptres": "scepters",
+  "scrutinise": "scrutinize",
+  "scrutinised": "scrutinized",
+  "scrutinises": "scrutinizes",
+  "scrutinising": "scrutinizing",
+  "secularisation": "secularization",
+  "secularise": "secularize",
+  "secularised": "secularized",
+  "secularises": "secularizes",
+  "secularising": "secularizing",
+  "sensationalise": "sensationalize",
+  "sensationalised": "sensationalized",
+  "sensationalises": "sensationalizes",
+  "sensationalising": "sensationalizing",
+  "sensitise": "sensitize",
+  "sensitised": "sensitized",
+  "sensitises": "sensitizes",
+  "sensitising": "sensitizing",
+  "sentimentalise": "sentimentalize",
+  "sentimentalised": "sentimentalized",
+  "sentimentalises": "sentimentalizes",
+  "sentimentalising": "sentimentalizing",
+  "sepulchre": "sepulcher",
+  "sepulchres": "sepulchers",
+  "serialisation": "serialization",
+  "serialisations": "serializations",
+  "serialise": "serialize",
+  "serialised": "serialized",
+  "serialises": "serializes",
+  "serialising": "serializing",
+  "sermonise": "sermonize",
+  "sermonised": "sermonized",
+  "sermonises": "sermonizes",
+  "sermonising": "sermonizing",
+  "sheikh": "sheik",
+  "shovelled": "shoveled",
+  "shovelling": "shoveling",
+  "shrivelled": "shriveled",
+  "shrivelling": "shriveling",
+  "signalise": "signalize",
+  "signalised": "signalized",
+  "signalises": "signalizes",
+  "signalising": "signalizing",
+  "signalled": "signaled",
+  "signalling": "signaling",
+  "smoulder": "smolder",
+  "smouldered": "smoldered",
+  "smouldering": "smoldering",
+  "smoulders": "smolders",
+  "snivelled": "sniveled",
+  "snivelling": "sniveling",
+  "snorkelled": "snorkeled",
+  "snorkelling": "snorkeling",
+  "snowplough": "snowplow",
+  "snowploughs": "snowplow",
+  "socialisation": "socialization",
+  "socialise": "socialize",
+  "socialised": "socialized",
+  "socialises": "socializes",
+  "socialising": "socializing",
+  "sodomise": "sodomize",
+  "sodomised": "sodomized",
+  "sodomises": "sodomizes",
+  "sodomising": "sodomizing",
+  "solemnise": "solemnize",
+  "solemnised": "solemnized",
+  "solemnises": "solemnizes",
+  "solemnising": "solemnizing",
+  "sombre": "somber",
+  "specialisation": "specialization",
+  "specialisations": "specializations",
+  "specialise": "specialize",
+  "specialised": "specialized",
+  "specialises": "specializes",
+  "specialising": "specializing",
+  "spectre": "specter",
+  "spectres": "specters",
+  "spiralled": "spiraled",
+  "spiralling": "spiraling",
+  "splendour": "splendor",
+  "splendours": "splendors",
+  "squirrelled": "squirreled",
+  "squirrelling": "squirreling",
+  "stabilisation": "stabilization",
+  "stabilise": "stabilize",
+  "stabilised": "stabilized",
+  "stabiliser": "stabilizer",
+  "stabilisers": "stabilizers",
+  "stabilises": "stabilizes",
+  "stabilising": "stabilizing",
+  "standardisation": "standardization",
+  "standardise": "standardize",
+  "standardised": "standardized",
+  "standardises": "standardizes",
+  "standardising": "standardizing",
+  "stencilled": "stenciled",
+  "stencilling": "stenciling",
+  "sterilisation": "sterilization",
+  "sterilisations": "sterilizations",
+  "sterilise": "sterilize",
+  "sterilised": "sterilized",
+  "steriliser": "sterilizer",
+  "sterilisers": "sterilizers",
+  "sterilises": "sterilizes",
+  "sterilising": "sterilizing",
+  "stigmatisation": "stigmatization",
+  "stigmatise": "stigmatize",
+  "stigmatised": "stigmatized",
+  "stigmatises": "stigmatizes",
+  "stigmatising": "stigmatizing",
+  "storey": "story",
+  "storeys": "stories",
+  "subsidisation": "subsidization",
+  "subsidise": "subsidize",
+  "subsidised": "subsidized",
+  "subsidiser": "subsidizer",
+  "subsidisers": "subsidizers",
+  "subsidises": "subsidizes",
+  "subsidising": "subsidizing",
+  "succour": "succor",
+  "succoured": "succored",
+  "succouring": "succoring",
+  "succours": "succors",
+  "sulphate": "sulfate",
+  "sulphates": "sulfates",
+  "sulphide": "sulfide",
+  "sulphides": "sulfides",
+  "sulphur": "sulfur",
+  "sulphurous": "sulfurous",
+  "summarise": "summarize",
+  "summarised": "summarized",
+  "summarises": "summarizes",
+  "summarising": "summarizing",
+  "swivelled": "swiveled",
+  "swivelling": "swiveling",
+  "symbolise": "symbolize",
+  "symbolised": "symbolized",
+  "symbolises": "symbolizes",
+  "symbolising": "symbolizing",
+  "sympathise": "sympathize",
+  "sympathised": "sympathized",
+  "sympathiser": "sympathizer",
+  "sympathisers": "sympathizers",
+  "sympathises": "sympathizes",
+  "sympathising": "sympathizing",
+  "synchronisation": "synchronization",
+  "synchronise": "synchronize",
+  "synchronised": "synchronized",
+  "synchronises": "synchronizes",
+  "synchronising": "synchronizing",
+  "synthesise": "synthesize",
+  "synthesised": "synthesized",
+  "synthesiser": "synthesizer",
+  "synthesisers": "synthesizers",
+  "synthesises": "synthesizes",
+  "synthesising": "synthesizing",
+  "syphon": "siphon",
+  "syphoned": "siphoned",
+  "syphoning": "siphoning",
+  "syphons": "siphons",
+  "systematisation": "systematization",
+  "systematise": "systematize",
+  "systematised": "systematized",
+  "systematises": "systematizes",
+  "systematising": "systematizing",
+  "tantalise": "tantalize",
+  "tantalised": "tantalized",
+  "tantalises": "tantalizes",
+  "tantalising": "tantalizing",
+  "tantalisingly": "tantalizingly",
+  "tasselled": "tasseled",
+  "technicolour": "technicolor",
+  "temporise": "temporize",
+  "temporised": "temporized",
+  "temporises": "temporizes",
+  "temporising": "temporizing",
+  "tenderise": "tenderize",
+  "tenderised": "tenderized",
+  "tenderises": "tenderizes",
+  "tenderising": "tenderizing",
+  "terrorise": "terrorize",
+  "terrorised": "terrorized",
+  "terrorises": "terrorizes",
+  "terrorising": "terrorizing",
+  "theatre": "theater",
+  "theatregoer": "theatergoer",
+  "theatregoers": "theatergoers",
+  "theatres": "theaters",
+  "theorise": "theorize",
+  "theorised": "theorized",
+  "theorises": "theorizes",
+  "theorising": "theorizing",
+  "tonne": "ton",
+  "tonnes": "tons",
+  "towelled": "toweled",
+  "towelling": "toweling",
+  "toxaemia": "toxemia",
+  "tranquillise": "tranquilize",
+  "tranquillised": "tranquilized",
+  "tranquilliser": "tranquilizer",
+  "tranquillisers": "tranquilizers",
+  "tranquillises": "tranquilizes",
+  "tranquillising": "tranquilizing",
+  "tranquillity": "tranquility",
+  "tranquillize": "tranquilize",
+  "tranquillized": "tranquilized",
+  "tranquillizer": "tranquilizer",
+  "tranquillizers": "tranquilizers",
+  "tranquillizes": "tranquilizes",
+  "tranquillizing": "tranquilizing",
+  "tranquilly": "tranquility",
+  "transistorised": "transistorized",
+  "traumatise": "traumatize",
+  "traumatised": "traumatized",
+  "traumatises": "traumatizes",
+  "traumatising": "traumatizing",
+  "travelled": "traveled",
+  "traveller": "traveler",
+  "travellers": "travelers",
+  "travelling": "traveling",
+  "travelog": "travelogue",
+  "travelogs": "travelogues",
+  "trialled": "trialed",
+  "trialling": "trialing",
+  "tricolour": "tricolor",
+  "tricolours": "tricolors",
+  "trivialise": "trivialize",
+  "trivialised": "trivialized",
+  "trivialises": "trivializes",
+  "trivialising": "trivializing",
+  "tumour": "tumor",
+  "tumours": "tumors",
+  "tunnelled": "tunneled",
+  "tunnelling": "tunneling",
+  "tyrannise": "tyrannize",
+  "tyrannised": "tyrannized",
+  "tyrannises": "tyrannizes",
+  "tyrannising": "tyrannizing",
+  "tyre": "tire",
+  "tyres": "tires",
+  "unauthorised": "unauthorized",
+  "uncivilised": "uncivilized",
+  "underutilised": "underutilized",
+  "unequalled": "unequaled",
+  "unfavourable": "unfavorable",
+  "unfavourably": "unfavorably",
+  "unionisation": "unionization",
+  "unionise": "unionize",
+  "unionised": "unionized",
+  "unionises": "unionizes",
+  "unionising": "unionizing",
+  "unorganised": "unorganized",
+  "unravelled": "unraveled",
+  "unravelling": "unraveling",
+  "unrecognisable": "unrecognizable",
+  "unrecognised": "unrecognized",
+  "unrivalled": "unrivaled",
+  "unsavoury": "unsavory",
+  "untrammelled": "untrammeled",
+  "urbanisation": "urbanization",
+  "urbanise": "urbanize",
+  "urbanised": "urbanized",
+  "urbanises": "urbanizes",
+  "urbanising": "urbanizing",
+  "utilisable": "utilizable",
+  "utilisation": "utilization",
+  "utilise": "utilize",
+  "utilised": "utilized",
+  "utilises": "utilizes",
+  "utilising": "utilizing",
+  "valour": "valor",
+  "vandalise": "vandalize",
+  "vandalised": "vandalized",
+  "vandalises": "vandalizes",
+  "vandalising": "vandalizing",
+  "vaporisation": "vaporization",
+  "vaporise": "vaporize",
+  "vaporised": "vaporized",
+  "vaporises": "vaporizes",
+  "vaporising": "vaporizing",
+  "vapour": "vapor",
+  "vapours": "vapors",
+  "verbalise": "verbalize",
+  "verbalised": "verbalized",
+  "verbalises": "verbalizes",
+  "verbalising": "verbalizing",
+  "victimisation": "victimization",
+  "victimise": "victimize",
+  "victimised": "victimized",
+  "victimises": "victimizes",
+  "victimising": "victimizing",
+  "videodisc": "videodisk",
+  "videodiscs": "videodisks",
+  "vigour": "vigor",
+  "visualisation": "visualization",
+  "visualisations": "visualizations",
+  "visualise": "visualize",
+  "visualised": "visualized",
+  "visualises": "visualizes",
+  "visualising": "visualizing",
+  "vocalisation": "vocalization",
+  "vocalisations": "vocalizations",
+  "vocalise": "vocalize",
+  "vocalised": "vocalized",
+  "vocalises": "vocalizes",
+  "vocalising": "vocalizing",
+  "vulcanised": "vulcanized",
+  "vulgarisation": "vulgarization",
+  "vulgarise": "vulgarize",
+  "vulgarised": "vulgarized",
+  "vulgarises": "vulgarizes",
+  "vulgarising": "vulgarizing",
+  "waggon": "wagon",
+  "waggons": "wagons",
+  "watercolour": "watercolor",
+  "watercolours": "watercolors",
+  "weaselled": "weaseled",
+  "weaselling": "weaseling",
+  "westernisation": "westernization",
+  "westernise": "westernize",
+  "westernised": "westernized",
+  "westernises": "westernizes",
+  "westernising": "westernizing",
+  "womanise": "womanize",
+  "womanised": "womanized",
+  "womaniser": "womanizer",
+  "womanisers": "womanizers",
+  "womanises": "womanizes",
+  "womanising": "womanizing",
+  "woollen": "woolen",
+  "woollens": "woolens",
+  "woollies": "woolies",
+  "woolly": "wooly",
+  "worshipped": "worshiped",
+  "worshipper": "worshiper",
+  "worshipping": "worshiping",
+  "yodelled": "yodeled",
+  "yodelling": "yodeling",
+  "yoghourt": "yogurt",
+  "yoghourts": "yogurts",
+  "yoghurt": "yogurt",
+  "yoghurts": "yogurts"
+}

pred.txt

@@ -0,0 +1 @@
+pred_str=['心 理 心 白 ', '血 燒 尻 川 ', '打 鼠 ', '十 三 項 ', '水 里 鄉 ', '這 擔 有 一 个 大 上 場 ', '聯 合 國 ', '菜 頭 人 仔 ', '賴 州 ', '肉 羹 ', '新 興 區 ', '刣 姑 平 ', '鯉 東 ', '推 仔 ', '西 園 邦 ', '心 喙 ', '燒 包 仔 ', '康 頭 ', '鳳 坑 ', '羅 水 溪 出 代 誌 ', '粗 勢 ', '哀 哀 ', '九 年 一 關 ', '精 靈 ', '三 寶 ', '索 索 ', '坐 船 咧 等 ', '眼 目 袂 曉 看 懸 低 ', '圳 毋 ', '兩 个 人 已 經 講 好 矣 ', '足 水 ', '細 空 喙 ', '你 仔 有 經 頭 ', '中 新 灣 ', '解 率 ', '草 人 坑 ', '祝 你 生 日 快 樂 ', '坑 仔 縫 ', '出 水 ', '皮 平 串 ', '耳 仔 街 ', '活 躼 ', '臺 灣 人 ', '落 崙 ', '婉 里 坑 ', '坑 仔 內 ', '起 起 落 落 ', '戶 錢 ', '紅 蟲 ', '大 湖 ', '拋 拋 轎 ', '柴 犬 ', '大 位 王 ', '大 著 鼎 ', '頭 鑢 ', '鬆 餅 ', '喙 掩 籠 ', '龍 雲 ', '施 局 ', '縮 時 ', '煞 尾 ', '無 話 講 甲 好 ', '下 暗 磅 ', '一 籠 ', '一 定 哭 未 著 ', '看 一 个 影 生 一 个 囝 ', '倒 勢 ', '北 京 ', '規 陣 俗 姓 ', '肚 臟 ', '變 天 ', '一 工 九 甲 ', '混 沌 ', '有 規 項 菜 ', '厭 武 ', '六 角 區 ', '過 路 山 ', '予 袂 予 人 徛 ', '粗 暑 ', '這 咧 活 蠐 錢 ', '車 仔 ', '看 甲 清 清 粗 粗 ', '楞 神 ', '化 身 ', '白 囡 仔 ', '心 花 東 區 ', '番 鼎 仔 ', '流 擴 擴 ', '大 雞 血 ', '贏 城 真 頭 ', '閃 爬 肉 ', '骨 ', '龜 大 ', '水 雞 皮 仔 ', '夯 硬 糕 ', '抾 無 三 文 本 著 想 欲 做 土 公 ', '阿 姆 姨 仔 ', '臺 北 橋 ', '雲 林 縣 ', '臺 中 港 ', '水 理 ', '臺 南 市 ', '七 堵 ', '做 公 德 ', '臺 川 鄉 ', '青 仔 百 ', '消 ', '漢 臭 真 好 ', '餐 飲 ', '按 呢 生 ', '海 尾 ', '無 米 可 拄 著 卵 粿 ', '旗 鄉 小 細 ', '睏 ', '頭 尾 ', '五 斷 時 鐘 ', '政 商 ', '吐 蟯 咧 ', '動 物 性 膠 質 ', '數 文 精 ', '悠 悠 卡 ', '你 攏 無 效 睏 喔 ', '埤 頭 鄉 ', '信 義 大 安 路 口 ', '三 三 八 八 ', '你 有 的 毋 通 予 別 講 ', '白 寮 鄉 ', '你 莫 著 定 我 肚 臍 ', '新 園 路 ', '金 錢 毋 是 萬 鈴 的 ', '飛 鈴 仔 菜 ', '烰 料 理 ', '你 是 誰 ', '番 烏 ', '新 街 ', '合 同 ', '挺 挺 ', '番 麗 殼 ', '天 氣 冷 仔 你 著 愛 雞 踢 一 領 三 ', '我 到 底 看 著 啥 物 ', '手 番 仔 ', '姑 不 二 三 張 ', '天 格 ', '透 珍 仔 ', '漚 叫 聲 ', '隔 音 白 ', '拍 通 頭 ', '開 花 滿 天 芳 結 子 遮 夯 人 ', '東 北 貴 風 ', '無 人 機 ', '做 牛 無 小 栗 ', '真 空 排 放 ', '這 搖 仔 活 落 去 真 緊 著 的 替 風 仔 ', '毛 瑞 仔 ', '我 喙 歹 ', '陽 明 山 ', '咖 啡 館 ', '囡 仔 食 ', '佇 對 流 針 一 點 點 ', '伊 規 心 欲 苦 大 學 ', '東 湖 國 中 ', '中 原 公 園 ', '半 路 店 ', '屏 南 坑 ', '東 石 鄉 ', '拚 拚 哨 的 ', '無 中 心 化 ', '肉 鬃 角 ', '典 型 氣 ', '鼠 邊 頭 尾 ', '波 丸 街 ', '人 愛 烘 粥 路 口 ', '千 葉 ', '切 鑠 度 ', '帶 擊 帶 離 ', '尪 仔 頭 囝 ', '飼 ', '咱 做 代 誌 的 時 圳 ', '包 水 轎 ', '海 灶 毋 知 尾 海 柴 破 袂 開 ', '遠 總 場 ', '健 康 中 生 ', '印 圇 醫 師 仔 ', '心 理 族 ', '獅 仔 頭 ', '大 肚 路 ', '飼 仔 ', '鹿 爺 區 ', '學 子 ', '放 懶 山 ', '虎 尾 竹 ', '卸 腹 肚 ', '興 順 鎮 ', '小 血 球 ', '化 格 ', '我 三 攏 輕 鬆 掛 領 ', '溜 落 來 ', '半 條 命 ', '心 我 習 慣 ', '陳 影 弓 ', '菜 寶 卵 ', '土 塗 塗 ', '鐵 甲 巫 ', '風 火 頭 ', '現 流 仔 ', '白 鑢 公 ', '悾 港 ', '較 大 ', '公 眾 橋 ', '巴 結 ', '果 汁 ', '有 拜 無 一 好 ', '這 時 ', '竹 百 ', '下 晝 ', '較 早 睏 ', '寮 寮 帶 順 ', '公 司 ', '看 看 緊 緊 ', '坑 仔 ', '若 使 香 菇 ', '山 貓 ', '大 頭 猴 母 ', '塗 板 人 豆 ', '奶 ', '內 延 ', '滾 下 鄉 林 ', '推 磨 ', '真 有 喙 水 ', '摩 托 斯 ', '想 想 眼 你 就 是 你 朋 友 ', '青 山 王 ', '澳 抾 蘭 ', '金 東 ', '便 仔 ', '陳 奕 債 ', '九 尾 金 ', '相 空 ', '過 東 跤 ', '頭 順 傾 傾 ', '某 定 著 ', '礦 機 仔 ', '跙 湧 ', '中 州 ', '青 天 街 ', '東 方 ', '文 底 ', '都 原 性 別 ', '舊 廁 ', '貓 潲 ', '法 人 國 小 ', '點 破 ', '公 臺 ', '某 起 姑 仔 ', '蠟 豆 ', '中 環 ', '鑢 死 瓜 活 ', '打 著 歹 對 步 ', '挑 俘 ', '夠 水 公 的 無 尊 姓 ', '交 力 片 ', '如 此 如 此 嚴 重 ', '一 个 一 个 ', '陳 港 大 橋 ', '上 大 代 誌 說 ', '瑞 仔 旭 光 ', '茴 東 跤 ', '旭 園 湖 ', '講 啥 物 烏 仔 漢 字 ', '爺 ', '刺 鼠 菇 毒 著 蟲 ', '袂 使 毋 知 臭 ', '斗 南 ', '姑 哺 笑 無 事 ', '水 粿 仔 ', '滿 狀 ', '手 礦 ', '心 腹 無 厭 ', '敢 食 戛 油 ', '阿 寶 磚 ', '機 動 戰 士 ', '變 動 ', '興 一 病 母 ', '敲 仙 鼓 ', '所 羅 門 君 堵 ', '終 界 ', '我 愛 臺 語 ', '校 鼎 ', '甜 甜 ', '上 蓋 ', '海 �� ', '落 葉 不 絲 ', '踏 車 巫 ', '殕 仔 徛 豬 仔 ', '方 琴 尪 仔 ', '山 ', '生 甲 袂 䆀 ', '九 尾 素 ', '倚 刀 仔 ', '中 央 大 學 ', '石 門 水 庫 ', '暖 手 相 ', '百 貨 店 ', '倯 素 甘 ', '較 蓋 ', '洗 身 仔 枋 ', '天 台 ', '數 字 片 ', '漚 愷 ', '鈍 市 會 ', '頭 走 淚 ', '夭 壽 鬼 ', '好 所 在 ', '軟 斜 ', '撤 離 ', '現 辦 ', '究 勘 話 ', '伊 心 做 遮 爾 擦 爾 ', '無 神 那 縱 深 ', '一 點 點 仔 ', '龍 江 ', '隔 查 某 囝 ', '推 著 ', '心 心 酒 店 ', '過 死 直 死 ', '較 平 咧 ', '透 生 大 汗 ', '漉 糊 白 仔 ', '行 避 址 袂 著 愛 人 生 ', '雞 花 仔 ', '那 三 年 仔 ', '荷 里 雕 ', '石 板 ', '虎 魚 ', '落 雨 了 喔 頭 殼 陂 間 擔 落 落 ', '較 聳 ', '雞 人 飼 ', '你 欲 行 對 佗 位 去 ', '敦 化 胡 平 路 口 ', '港 仔 喙 ', '我 影 今 年 二 十 出 頭 矣 ', '心 肝 醫 ', '下 晝 覺 武 ', '一 分 ', '蠓 蠓 仔 蠓 ', '完 安 ', '牛 郎 ', '袂 看 得 ', '立 公 立 直 ', '龜 山 島 ', '過 渡 ', '塑 做 堆 ', '尾 仔 跤 ', '包 大 ', '原 地 ', '三 頂 三 重 ', '林 仔 口 ', '畢 電 ', '草 拗 ', '揣 轉 愛 ', '機 中 站 ', '苴 仔 代 ', '心 酺 ', '燒 桌 ', '頂 底 寮 ', '竹 南 ', '你 共 包 仔 提 起 嚇 小 一 个 ', '平 方 筋 平 方 筋 ', '頂 公 館 ', '線 仙 ', '演 藝 界 ', '血 本 ', '毋 知 無 怨 ', '較 緊 ', '有 也 好 無 也 好 ', '草 鑠 ', '講 推 仔 龍 ', '怪 奇 菇 ', '國 抱 加 盟 ', '新 羅 ', '過 去 ', '豆 豆 仔 ', '拍 寶 ', '物 件 若 予 伊 提 過 手 ', '奶 仔 嬰 ', '意 蓋 ', '刺 尾 風 ', '烏 來 ', '挺 走 ', '無 踏 無 踅 ', '我 袂 準 備 落 山 ', '倒 擁 ', '聽 無 ', '喜 的 頭 雞 娘 對 人 家 正 好 咧 ', '大 細 輪 ', '五 葉 ', '無 年 五 你 會 出 國 讀 冊 ', '借 園 ', '虼 兒 ', '後 寮 ', '我 攏 有 看 著 ', '建 人 路 ', '歹 天 ', '三 跤 步 一 坎 唸 ', '賴 肉 ', '龜 你 祝 ', '下 晡 點 心 ', '冷 管 ', '無 枝 結 ', '公 文 ', '雜 喙 狗 尻 川 ', '頭 爺 屁 ', '橐 錢 ', '茂 空 石 ', '蚵 仔 雞 湯 仔 ', '雞 使 面 ', '白 對 頭 ', '蜂 一 枝 ', '無 一 步 ', '路 口 講 出 來 ', '頭 頂 ', '防 地 做 甲 流 汗 防 跤 相 甲 流 瀾 ', '好 鬥 陣 ', '拍 死 板 ', '同 陂 ', '打 巴 ', '簡 順 ', '一 个 才 一 个 ', '緊 來 選 ', '洋 仔 話 ', '拍 數 ', '你 拉 瓦 ', '總 舖 ', '山 里 無 間 龜 溜 溜 ', '鑿 瓜 ', '手 心 動 ', '電 子 ', '暗 空 ', '散 食 ', '塑 膠 車 ', '要 意 ', '燒 雞 胿 ', '吊 猴 ', '掠 猴 ', '暗 頭 仔 ', '濟 囝 濟 八 八 濟 新 婦 濟 體 踢 ', '猴 球 ', '無 拄 好 ', '草 頭 貓 ', '金 寬 失 立 ', '觀 音 臺 師 ', '財 圳 ', '壥 暗 仔 ', '拗 路 ', '結 禮 ', '一 目 仔 ', '提 款 片 ', '無 你 的 代 ', '天 公 債 ', '篡 廣 ', '離 櫥 ', '馬 鈴 薯 條 ', '做 塗 水 的 ', '林 爭 味 ', '流 明 傳 ', '官 包 油 ', '湯 醋 ', '離 間 ', '拍 鐵 獅 ', '萬 一 跤 步 ', '鐵 耙 的 ', '下 禮 拜 ', '死 亡 誌 悌 ', '你 毋 通 放 在 心 肝 內 ', '臭 牛 仔 ', '趁 一 空 食 三 冬 ', '我 欲 下 代 語 ', '臺 ', '拉 甲 斗 ', '雨 水 罨 ', '貴 州 姓 ', '敦 華 北 路 ', '你 看 起 來 是 正 年 仔 媠 ', '真 硬 食 公 公 硬 食 天 公 ', '回 去 ', '下 晡 時 ', '幾 个 ', '紅 燈 ', '感 情 ', '二 條 仔 ', '無 鐵 仔 ', '延 興 路 ', '馬 空 ', '六 點 ', '死 中 ', '超 寶 奶 ', '園 仔 機 ', '廣 寮 區 ', '甘 血 噴 天 ', '大 好 夯 人 ', '行 政 院 ', '食 飼 ', '狗 弟 厝 ', '冷 清 ', '我 的 港 ', '浸 水 器 ', '雞 柳 ', '天 公 山 ', '關 ', '愛 代 語 ', '我 共 欲 袂 予 ', '天 飯 ', '有 冷 汗 ', '衝 ', '中 國 話 ', '爿 香 對 排 ', '四 塊 草 ', '拆 箍 ', '關 湳 ', '結 霜 卵 ', '擔 貂 ', '大 公 山 ', '死 瓜 丸 大 平 ', '有 這 款 的 代 誌 我 攏 毋 知 ', '埔 頭 仔 機 ', '國 姓 爺 ', '那 口 區 ', '盤 ', '苦 山 林 斯 ', '代 誌 大 案 ', '半 姓 半 姨 ', '點 滴 點 滴 桃 坑 鄉 魑 施 ', '你 有 工 工 ', '任 通 無 洗 手 的 朋 友 ', '烏 日 ', '衛 風 ', '水 園 ', '怨 恬 ', '連 半 个 人 影 無 ', '吊 褲 ', '塗 子 工 ', '霧 沙 沙 ', '衛 生 箸 ', '鯉 丸 仔 ', '搵 醬 ', '你 老 母 拄 咧 掠 咧 ', '友 情 濟 來 順 ', '港 仔 喙 國 小 ', '漉 糊 塑 ', '落 尾 的 ', '平 方 電 腦 ', '蠓 蠓 仔 鰗 鰆 塗 嘛 會 擔 ', '冷 滾 水 ', '一 个 人 口 按 一 个 位 ', '兼 數 家 ', '雞 卵 糕 仔 ', '溼 灌 ', '珠 倒 ', '聖 誕 老 阿 公 ', '期 頂 白 毛 毛 ', '飯 碗 ', '南 縣 ', '毋 人 事 ', '餘 韻 ', '包 山 包 海 ', '兼 空 區 ', '洗 澡 精 ', '拋 勁 電 傷 ', '開 喙 涵 粒 粒 臭 ', '踏 水 器 ', '老 姑 精 ', '塑 膠 路 仔 ', '巣 位 仔 聯 名 ', '臭 水 ', '塑 膠 矸 仔 ', '衛 生 機 ', '面 開 ', '圍 場 天 下 不 亂 ', '錢 票 ', '郭 仔 樹 ', '牛 角 三 件 ', '半 夜 ', '思 雷 ', '衝 湧 ', '感 有 文 德 ', '食 涼 ', '有 的 紙 欠 濁 ', '做 塗 水 ', '怪 奇 ', '半 斤 ', '交 關 ', '海 岸 ', '中 山 路 ', '番 仔 狗 ', '婉 勇 仔 ', '龜 龜 毛 毛 ', '毋 捌 意 看 過 時 ', '標 段 ', '機 公 號 ', '海 拄 仔 ', '猴 頭 猴 ', '彼 死 兩 相 ', '民 主 ', '底 香 卵 ', '無 底 交 電 ', '食 人 一 口 還 人 一 斗 ', '鐵 段 ', '東 埤 仔 ', '年 猴 ', '佛 心 投 資 者 ', '人 講 對 人 講 ', '金 剛 戰 事 ', '做 肉 餅 ', '小 補 喙 ', '櫻 桮 魚 ', '美 頭 船 ', '一 筆 錢 ', '割 菜 ', '狗 馬 日 晟 ', '一 坵 草 一 點 鹿 ', '閘 時 ', '大 頭 母 人 ', '包 圍 ', '頂 哩 擺 ', '柔 珠 仔 ', '人 間 幾 幾 個 人 ', '無 氣 體 情 ', '日 落 山 ', '偏 鄉 ', '七 頭 跤 ', '袂 死 心 ', '戒 罪 ', '走 友 人 ', '有 聽 著 無 ', '大 湯 埕 ', '斗 人 仔 本 ', '滾 鼓 ', '影 人 ', '抢 查 某 ', '暗 當 食 聖 ', '草 鰆 鰡 仔 ', '痛 快 ', '手 包 ', '包 餅 ', '手 轎 ', '暖 無 ', '無 精 彩 ', '內 街 ', '烏 手 ', '燒 身 厝 ', '龜 潤 ', '五 公 厝 貯 ', '山 巡 ', '人 生 欠 貨 ', '天 公 無 大 著 ', '望 安 鄉 ', '食 虎 ', '分 子 ', '批 車 ', '大 支 ', '公 館 路 ', '碧 潭 路 ', '臺 北 後 酒 後 酒 小 學 步 ', '青 年 不 文 ', '四 分 仔 ', '頭 哭 的 ', '南 機 場 ', '湧 瑞 街 ', '浮 珠 溪 ', '池 上 ', '跤 車 ', '關 胎 ', '苦 袂 冬 ', '單 蒼 公 ', '睏 鏡 花 ', '車 驅 ', '共 人 嗙 ', '頂 底 園 ', '潮 州 街 口 ', '車 坤 ', '龍 頂 厝 ', '芳 樂 公 園 ', '金 剛 鎮 ', '忠 孝 福 興 ', '民 族 保 溫 路 口 ', '松 信 路 ', '糋 麭 粉 ', '文 州 ', '臺 北 蘭 路 山 ', '有 夠 咧 加 速 ', '伊 雞 我 一 擲 ', '笙 仔 ', '魚 豬 鄉 ', '囝 柑 仔 囝 ', '四 跤 水 仙 ', '暖 暖 區 ', '用 起 美 若 洗 喙 時 愛 上 如 此 高 ', '新 社 ', '武 神 力 故 宮 好 ', '實 包 ', '埔 尾 ', '寶 路 ', '冰 糖 ', '福 星 橋 ', '阿 拉 ', '聖 安 ', '遠 林 市 ', '歹 舌 ', '講 代 誌 ', '年 度 年 ', '景 興 路 ', '海 蛤 仔 頭 ', '一 稈 半 稈 ', '死 輪 節 ', '包 殼 ', '瑞 穂 ', '咱 思 貓 頭 鳥 ', '茂 羅 開 ', '酒 光 喙 罨 ', '果 汁 體 ', '蛤 術 ', '不 知 ', '遠 東 區 ', '上 海 茂 荷 ', '意 識 面 ', '光 芳 ', '話 語 指 中 ', '皮 人 積 肉 糜 ', '紅 豆 仔 麭 ', '頂 花 ', '相 伨 ', '三 兩 人 講 四 斤 話 ', '禁 欲 ', '一 坵 ', '卵 箍 ', '真 珠 看 做 貓 齒 使 ', '天 津 ', '角 運 動 龍 ', '莫 門 足 恐 怖 ', '品 德 街 ', '羅 典 ', '錢 拆 藥 ', '用 湯 平 ', '花 生 園 ', '可 睏 丸 ', '武 松 ', '豬 頭 鳳 ', '蔭 出 ', '閘 位 ', '敲 蟲 ', '番 薯 片 ', '摒 氣 ', '卵 子 ', '番 喜 甘 丸 ', '大 海 毋 驚 大 水 ', '橫 的 ', '跤 手 指 ', '歇 ', '千 巴 總 ', '徛 倒 規 ', '查 某 影 ', '番 薯 球 ', '蜂 寮 草 ', '大 火 ', '火 記 母 ', '英 國 話 ', '免 滴 顧 問 ', '老 街 ', '看 有 ', '摒 車 天 花 肚 ', '下 文 活 東 ', '臺 羅 姿 ', '內 ', '金 風 山 ', '死 袂 見 效 ', '嚇 死 ', '交 法 ', '馬 鈴 薯 片 ', '里 頭 拍 甲 蒼 仔 股 毋 戛 ', '營 養 庫 本 ', '放 水 機 ', '大 大 塊 仔 ', '宜 蘭 ', '無 一 定 敢 仔 我 看 著 ', '有 困 山 ', '代 誌 好 生 仔 尾 ', '媠 氣 ', '景 文 街 ', '酒 天 酒 地 ', '今 年 是 一 个 好 年 冬 ', '師 佮 ', '烏 山 頭 ', '三 講 四 毋 著 ', '武 術 文 字 ', '反 策 ', '大 日 本 地 國 ', '仝 蓋 網 上 ', '若 後 ', '吞 奶 ', '莫 閣 ', '緊 急 做 代 一 跤 要 緊 ', '你 講 啥 ', '經 水 ', '位 格 ', '生 體 健 康 ', '跤 手 十 足 ', '電 子 學 ', '夏 年 ', '大 路 大 股 ', '伊 是 一 个 大 米 奶 ', '解 油 ', '膨 氣 ', '你 陪 我 去 那 邊 好 無 ', '網 路 用 較 著 咧 ', '周 賽 ', '姑 誕 一 生 ', '家 己 ', '死 中 袂 終 ', '牛 舟 仔 ', '姑 烘 毒 黒 ', '離 蜢 ', '注 目 ', '掠 仔 ', '燒 桌 ', '啉 喙 ', '創 辦 人 ', '水 灿 林 ', '市 場 動 物 ', '目 睭 白 金 ', '抾 囡 仔 ', '我 予 滾 水 通 一 个 泡 泡 就 聽 的 ', '順 桮 ', '話 人 ', '郭 牙 ', '三 里 經 ', '飯 盒 ', '天 車 ', '圍 堵 ', '茂 仔 久 ', '白 蟳 ', '電 蜜 ', '鹽 酸 草 ', '番 仔 番 薯 ', '年 一 擺 來 你 攏 趨 好 了 的 ', '來 喔 緊 愛 看 喔 ', '番 界 ', '總 事 件 ', '破 柴 ', '列 位 ', '鳳 山 ', '老 厝 ', '開 山 母 鄉 ', '金 米 國 小 ', '跌 腹 市 ', '橋 頭 區 ', '南 州 ', '好 米 鎮 ', '中 立 ', '西 區 ', '國 泰 ', '水 轉 流 ', '中 央 路 ', '南 福 著 ', '金 華 大 哈 ', '使 甲 ', '士 林 區 ', '紙 票 變 厚 ', '中 山 橋 ', '彰 化 人 烘 ', '二 國 倉 ', '喙 風 轎 ', '花 仔 菜 ', '中 間 國 小 ', '按 呢 著 好 ', '老 去 矣 ', '車 埕 ', '南 洋 街 ', '匈 牙 ', '松 壽 路 口 ', '中 坑 ', '不 屬 志 溪 ', '瑞 源 ', '後 驛 ', '大 �� 鼎 ', '鳥 仔 弓 ', '再 掰 ', '新 園 區 ', '政 義 遊 腳 ', '寧 夏 路 ', '動 蠻 ', '大 好 蜈 蚣 ', '仁 愛 鄉 ', '看 龍 ', '班 ', '西 米 粉 ', '食 機 ', '吊 管 街 ', '仙 師 娘 ', '拍 鐵 寮 ', '路 影 區 ', '韻 尾 ', '倚 雙 溪 ', '信 義 區 ', '你 母 甘 知 仔 你 一 隻 發 肺 文 ', '敢 按 呢 ', '狗 草 番 ', '望 星 橋 ', '肌 膚 ', '擦 耳 膠 ', '文 聳 橋 ', '有 要 緊 無 ', '若 來 鄉 ', '紅 目 石 獅 ', '囡 仔 人 嘛 看 有 ', '青 大 ', '文 鼓 炊 ', '歹 喙 道 ', '查 某 人 ', '魯 菜 頭 ', '跳 尾 ', '鴨 母 大 正 堆 ', '不 貪 不 素 ', '徛 名 ', '白 目 粉 ', '聽 阿 毛 講 場 ', '豆 磨 ', '婉 明 府 ', '粗 花 ', '雞 人 蜂 ', '暗 眠 膜 山 貓 ', '袂 曉 講 話 ', '遊 蠢 ', '巡 事 ', '三 攏 無 ', '鐵 葉 膏 ', '推 桌 ', '足 夠 無 看 的 ', '起 雞 母 皮 ', '臺 港 ', '北 建 圈 ', '爛 花 踏 ', '著 瓦 場 ', '苦 寮 仔 ', '五 起 肩 頭 仔 交 起 來 的 達 基 馬 騰 ', '絨 帽 位 ', '厝 尾 頂 ', '花 樹 人 ', '大 箍 ', '較 低 ', '客 戶 ', '鴨 母 燒 ', '烏 仔 箍 ', '海 口 ', '阿 里 不 達 ', '冰 島 ', '節 ', '使 啦 ', '臭 使 星 ', '網 網 爺 尻 川 ', '冊 鑿 ', '陽 山 迴 腸 ', '你 冊 螺 絲 格 ', '小 山 郎 仔 ', '剪 仔 菇 ', '青 梅 仔 目 鏡 ', '真 愛 ', '在 地 大 學 ', '欲 去 佗 位 ', '崁 頭 鴨 ', '免 流 ', '揣 著 ', '橋 頭 走 路 ', '門 東 仔 ', '冷 角 龍 ', '狗 頭 殕 容 ', '傳 番 ', '我 欲 去 變 數 ', '那 頓 食 ', '鼻 雷 公 ', '代 誌 文 好 ', '抌 頭 ', '牛 牙 使 彈 ', '材 料 科 學 ', '掩 地 水 街 ', '貨 錢 ', '睏 袂 去 ']

preprocessor_config.json

@@ -0,0 +1,14 @@
+{
+  "chunk_length": 30,
+  "feature_extractor_type": "WhisperFeatureExtractor",
+  "feature_size": 80,
+  "hop_length": 160,
+  "n_fft": 400,
+  "n_samples": 480000,
+  "nb_max_frames": 3000,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "WhisperProcessor",
+  "return_attention_mask": false,
+  "sampling_rate": 16000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ba901ed975c5cd4ccdcf22f79d75ab9211d9b2a28135620fd264d554d5f2684
+size 3055754841

run.sh

@@ -0,0 +1,34 @@
+python3 run_speech_recognition_seq2seq_streaming.py \
+--model_name_or_path="openai/whisper-medium" \
+--dataset_name="mozilla-foundation/common_voice_11_0" \
+--dataset_config_name="nan-tw" \
+--language="chinese" \
+--train_split_name="train" \
+--eval_split_name="test" \
+--model_index_name="Whisper medium nan-tw" \
+--max_steps="5000" \
+--output_dir="./" \
+--per_device_train_batch_size="2" \
+--per_device_eval_batch_size="2" \
+--logging_steps="25" \
+--learning_rate="1e-5" \
+--warmup_steps="500" \
+--evaluation_strategy="steps" \
+--eval_steps="1000" \
+--save_strategy="steps" \
+--save_steps="1000" \
+--generation_max_length="225" \
+--length_column_name="input_length" \
+--max_duration_in_seconds="30" \
+--text_column_name="sentence" \
+--freeze_feature_encoder="False" \
+--report_to="tensorboard" \
+--gradient_checkpointing \
+--fp16 \
+--overwrite_output_dir \
+--do_train \
+--do_eval \
+--predict_with_generate \
+--do_normalize_eval \
+--use_auth_token \
+--optim="adamw_bnb_8bit" 

run_speech_recognition_seq2seq_streaming.py

@@ -0,0 +1,617 @@
+#!/usr/bin/env python
+# coding=utf-8
+# Copyright 2022 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+Fine-tuning the library models for sequence to sequence speech recognition
+with 🤗 Datasets' streaming mode.
+"""
+# You can also adapt this script for your own sequence to sequence speech
+# recognition task. Pointers for this are left as comments.
+import ryNormText
+import logging
+import os
+import re
+import string
+import sys
+from dataclasses import dataclass, field
+from typing import Any, Dict, List, Optional, Union
+
+import datasets
+import torch
+from datasets import IterableDatasetDict, interleave_datasets, load_dataset
+from torch.utils.data import IterableDataset
+
+import evaluate
+import transformers
+from transformers import (
+    AutoConfig,
+    AutoFeatureExtractor,
+    AutoModelForSpeechSeq2Seq,
+    AutoProcessor,
+    AutoTokenizer,
+    HfArgumentParser,
+    Seq2SeqTrainer,
+    Seq2SeqTrainingArguments,
+    TrainerCallback,
+    set_seed,
+)
+from transformers.trainer_pt_utils import IterableDatasetShard
+from transformers.trainer_utils import get_last_checkpoint, is_main_process
+from transformers.utils import check_min_version, send_example_telemetry
+from transformers.utils.versions import require_version
+
+# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
+check_min_version("4.25.0.dev0")
+
+require_version("datasets>=1.18.2", "To fix: pip install -r examples/pytorch/speech-recognition/requirements.txt")
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class ModelArguments:
+    """
+    Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.
+    """
+
+    model_name_or_path: str = field(
+        metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"}
+    )
+    config_name: Optional[str] = field(
+        default=None, metadata={"help": "Pretrained config name or path if not the same as model_name"}
+    )
+    tokenizer_name: Optional[str] = field(
+        default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
+    )
+    feature_extractor_name: Optional[str] = field(
+        default=None, metadata={"help": "feature extractor name or path if not the same as model_name"}
+    )
+    cache_dir: Optional[str] = field(
+        default=None,
+        metadata={"help": "Where to store the pretrained models downloaded from huggingface.co"},
+    )
+    use_fast_tokenizer: bool = field(
+        default=True,
+        metadata={"help": "Whether to use one of the fast tokenizer (backed by the tokenizers library) or not."},
+    )
+    model_revision: str = field(
+        default="main",
+        metadata={"help": "The specific model version to use (can be a branch name, tag name or commit id)."},
+    )
+    use_auth_token: bool = field(
+        default=False,
+        metadata={
+            "help": (
+                "Will use the token generated when running `huggingface-cli login` (necessary to use this script "
+                "with private models)."
+            )
+        },
+    )
+    freeze_feature_encoder: bool = field(
+        default=True, metadata={"help": "Whether to freeze the feature encoder layers of the model."}
+    )
+    freeze_encoder: bool = field(
+        default=False, metadata={"help": "Whether to freeze the entire encoder of the seq2seq model."}
+    )
+    forced_decoder_ids: List[List[int]] = field(
+        default=None,
+        metadata={
+            "help": (
+                "A list of pairs of integers which indicates a mapping from generation indices to token indices "
+                "that will be forced before sampling. For example, [[0, 123]] means the first generated token "
+                "will always be a token of index 123."
+            )
+        },
+    )
+    suppress_tokens: List[int] = field(
+        default=None, metadata={"help": "A list of tokens that will be suppressed at generation."}
+    )
+    model_index_name: str = field(default=None, metadata={"help": "Pretty name for the model card."})
+
+
+@dataclass
+class DataTrainingArguments:
+    """
+    Arguments pertaining to what data we are going to input our model for training and eval.
+    """
+
+    dataset_name: str = field(
+        default=None, metadata={"help": "The name of the dataset to use (via the datasets library)."}
+    )
+    dataset_config_name: Optional[str] = field(
+        default=None, metadata={"help": "The configuration name of the dataset to use (via the datasets library)."}
+    )
+    text_column: Optional[str] = field(
+        default=None,
+        metadata={"help": "The name of the column in the datasets containing the full texts (for summarization)."},
+    )
+    max_train_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": (
+                "For debugging purposes or quicker training, truncate the number of training examples to this "
+                "value if set."
+            )
+        },
+    )
+    max_eval_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": (
+                "For debugging purposes or quicker training, truncate the number of evaluation examples to this "
+                "value if set."
+            )
+        },
+    )
+    audio_column_name: str = field(
+        default="audio",
+        metadata={"help": "The name of the dataset column containing the audio data. Defaults to 'audio'"},
+    )
+    text_column_name: str = field(
+        default="text",
+        metadata={"help": "The name of the dataset column containing the text data. Defaults to 'text'"},
+    )
+    max_duration_in_seconds: float = field(
+        default=20.0,
+        metadata={
+            "help": (
+                "Truncate audio files that are longer than `max_duration_in_seconds` seconds to"
+                " 'max_duration_in_seconds`"
+            )
+        },
+    )
+    min_duration_in_seconds: float = field(
+        default=0.0, metadata={"help": "Filter audio files that are shorter than `min_duration_in_seconds` seconds"}
+    )
+    train_split_name: str = field(
+        default="train",
+        metadata={
+            "help": "The name of the training data set split to use (via the datasets library). Defaults to 'train'"
+        },
+    )
+    eval_split_name: str = field(
+        default="test",
+        metadata={
+            "help": "The name of the training data set split to use (via the datasets library). Defaults to 'train'"
+        },
+    )
+    do_lower_case: bool = field(
+        default=False,
+        metadata={"help": "Whether the target text should be lower cased."},
+    )
+    do_remove_punctuation: bool = field(
+        default=False,
+        metadata={"help": "Whether the target text should be striped of punctuation."},
+    )
+    do_normalize_eval: bool = field(
+        default=True,
+        metadata={"help": "Whether to normalise the references and predictions in the eval WER calculation."},
+    )
+    language: str = field(
+        default=None,
+        metadata={
+            "help": (
+                "Language for multilingual fine-tuning. This argument should be set for multilingual fine-tuning "
+                "only. For English speech recognition, it should be set to `None`."
+            )
+        },
+    )
+    task: str = field(
+        default="transcribe",
+        metadata={"help": "Task, either `transcribe` for speech recognition or `translate` for speech translation."},
+    )
+    shuffle_buffer_size: Optional[int] = field(
+        default=500,
+        metadata={
+            "help": (
+                "The number of streamed examples to download before shuffling them. The large the buffer, "
+                "the closer it is to real offline shuffling."
+            )
+        },
+    )
+
+
+@dataclass
+class DataCollatorSpeechSeq2SeqWithPadding:
+    """
+    Data collator that will dynamically pad the inputs received.
+    Args:
+        processor ([`WhisperProcessor`])
+            The processor used for processing the data.
+        decoder_start_token_id (`int`)
+            The begin-of-sentence of the decoder.
+    """
+
+    processor: Any
+    decoder_start_token_id: int
+
+    def __call__(self, features: List[Dict[str, Union[List[int], torch.Tensor]]]) -> Dict[str, torch.Tensor]:
+        # split inputs and labels since they have to be of different lengths and need
+        # different padding methods
+        model_input_name = self.processor.model_input_names[0]
+        input_features = [{model_input_name: feature[model_input_name]} for feature in features]
+        label_features = [{"input_ids": feature["labels"]} for feature in features]
+
+        batch = self.processor.feature_extractor.pad(input_features, return_tensors="pt")
+
+        labels_batch = self.processor.tokenizer.pad(label_features, return_tensors="pt")
+
+        # replace padding with -100 to ignore loss correctly
+        labels = labels_batch["input_ids"].masked_fill(labels_batch.attention_mask.ne(1), -100)
+
+        # if bos token is appended in previous tokenization step,
+        # cut bos token here as it's append later anyways
+        if (labels[:, 0] == self.decoder_start_token_id).all().cpu().item():
+            labels = labels[:, 1:]
+
+        batch["labels"] = labels
+
+        return batch
+
+
+def load_streaming_dataset(dataset_name, dataset_config_name, split="train", **kwargs):
+    """
+    Utility function to load a dataset in streaming mode. For datasets with multiple splits,
+    each split is loaded individually and then splits combined by taking alternating examples from
+    each (interleaving).
+    """
+    if "+" in split:
+        # load multiple splits separated by the `+` symbol with streaming mode
+        dataset_splits = [
+            load_dataset(dataset_name, dataset_config_name, split=split_name, streaming=True, **kwargs)
+            for split_name in split.split("+")
+        ]
+        # interleave multiple splits to form one dataset
+        interleaved_dataset = interleave_datasets(dataset_splits)
+        return interleaved_dataset
+    else:
+        # load a single split *with* streaming mode
+        dataset = load_dataset(dataset_name, dataset_config_name, split=split, streaming=True, **kwargs)
+        return dataset
+
+
+def main():
+    # 1. Parse input arguments
+    # See all possible arguments in src/transformers/training_args.py
+    # or by passing the --help flag to this script.
+    # We now keep distinct sets of args, for a cleaner separation of concerns.
+    parser = HfArgumentParser((ModelArguments, DataTrainingArguments, Seq2SeqTrainingArguments))
+
+    if len(sys.argv) == 2 and sys.argv[1].endswith(".json"):
+        # If we pass only one argument to the script and it's the path to a json file,
+        # let's parse it to get our arguments.
+        model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))
+    else:
+        model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_speech_recognition_seq2seq_streaming", model_args, data_args)
+
+    # 2. Setup logging
+    logging.basicConfig(
+        format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
+        datefmt="%m/%d/%Y %H:%M:%S",
+        handlers=[logging.StreamHandler(sys.stdout)],
+    )
+    log_level = training_args.get_process_log_level()
+    logger.setLevel(log_level)
+    datasets.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.enable_default_handler()
+    transformers.utils.logging.enable_explicit_format()
+
+    logger.setLevel(logging.INFO if is_main_process(training_args.local_rank) else logging.WARN)
+
+    # Log on each process the small summary:
+    logger.warning(
+        f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
+        f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
+    )
+    logger.info(f"Training/evaluation parameters {training_args}")
+
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(training_args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+    logger.info("Training/evaluation parameters %s", training_args)
+
+    # 3. Detecting last checkpoint and eventually continue from last checkpoint
+    last_checkpoint = None
+    if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir:
+        last_checkpoint = get_last_checkpoint(training_args.output_dir)
+        if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0:
+            raise ValueError(
+                f"Output directory ({training_args.output_dir}) already exists and is not empty. "
+                "Use --overwrite_output_dir to overcome."
+            )
+        elif last_checkpoint is not None and training_args.resume_from_checkpoint is None:
+            logger.info(
+                f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change "
+                "the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
+            )
+
+    # Set seed before initializing model.
+    set_seed(training_args.seed)
+
+    # 4. Load dataset
+    raw_datasets = IterableDatasetDict()
+
+    if training_args.do_train:
+        raw_datasets["train"] = load_streaming_dataset(
+            data_args.dataset_name,
+            data_args.dataset_config_name,
+            split=data_args.train_split_name,
+            use_auth_token=True if model_args.use_auth_token else None,
+        )
+
+    if training_args.do_eval:
+        raw_datasets["eval"] = load_streaming_dataset(
+            data_args.dataset_name,
+            data_args.dataset_config_name,
+            split=data_args.eval_split_name,
+            use_auth_token=True if model_args.use_auth_token else None,
+        )
+
+    raw_datasets_features = list(next(iter(raw_datasets.values())).features.keys())
+
+    if data_args.audio_column_name not in raw_datasets_features:
+        raise ValueError(
+            f"--audio_column_name '{data_args.audio_column_name}' not found in dataset '{data_args.dataset_name}'. "
+            "Make sure to set `--audio_column_name` to the correct audio column - one of "
+            f"{', '.join(raw_datasets_features)}."
+        )
+
+    if data_args.text_column_name not in raw_datasets_features:
+        raise ValueError(
+            f"--text_column_name {data_args.text_column_name} not found in dataset '{data_args.dataset_name}'. "
+            "Make sure to set `--text_column_name` to the correct text column - one of "
+            f"{', '.join(raw_datasets_features)}."
+        )
+
+    # 5. Load pretrained model, tokenizer, and feature extractor
+    #
+    # Distributed training:
+    # The .from_pretrained methods guarantee that only one local process can concurrently
+    config = AutoConfig.from_pretrained(
+        model_args.config_name if model_args.config_name else model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        revision=model_args.model_revision,
+        use_auth_token=True if model_args.use_auth_token else None,
+    )
+
+    config.update({"forced_decoder_ids": model_args.forced_decoder_ids, "suppress_tokens": model_args.suppress_tokens})
+
+    feature_extractor = AutoFeatureExtractor.from_pretrained(
+        model_args.feature_extractor_name if model_args.feature_extractor_name else model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        revision=model_args.model_revision,
+        use_auth_token=True if model_args.use_auth_token else None,
+    )
+    tokenizer = AutoTokenizer.from_pretrained(
+        model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        use_fast=model_args.use_fast_tokenizer,
+        revision=model_args.model_revision,
+        use_auth_token=True if model_args.use_auth_token else None,
+    )
+    model = AutoModelForSpeechSeq2Seq.from_pretrained(
+        model_args.model_name_or_path,
+        config=config,
+        cache_dir=model_args.cache_dir,
+        revision=model_args.model_revision,
+        use_auth_token=True if model_args.use_auth_token else None,
+    )
+
+    if model.config.decoder_start_token_id is None:
+        raise ValueError("Make sure that `config.decoder_start_token_id` is correctly defined")
+
+    if model_args.freeze_feature_encoder:
+        model.freeze_feature_encoder()
+
+    if model_args.freeze_encoder:
+        model.freeze_encoder()
+        model.model.encoder.gradient_checkpointing = False
+
+    if data_args.language is not None:
+        # We only need to set the task id when the language is specified (i.e. in a multilingual setting)
+        tokenizer.set_prefix_tokens(language=data_args.language, task=data_args.task)
+
+    # 6. Resample speech dataset if necessary
+    dataset_sampling_rate = next(iter(raw_datasets.values())).features[data_args.audio_column_name].sampling_rate
+    if dataset_sampling_rate != feature_extractor.sampling_rate:
+        raw_datasets = raw_datasets.cast_column(
+            data_args.audio_column_name, datasets.features.Audio(sampling_rate=feature_extractor.sampling_rate)
+        )
+
+    # 7. Preprocessing the datasets.
+    # We need to read the audio files as arrays and tokenize the targets.
+    max_input_length = data_args.max_duration_in_seconds * feature_extractor.sampling_rate
+    min_input_length = data_args.min_duration_in_seconds * feature_extractor.sampling_rate
+    audio_column_name = data_args.audio_column_name
+    text_column_name = data_args.text_column_name
+    model_input_name = feature_extractor.model_input_names[0]
+    do_lower_case = data_args.do_lower_case
+    do_remove_punctuation = data_args.do_remove_punctuation
+
+    punctuation_to_remove = string.punctuation.replace("'", "")  # don't remove apostrophes
+    punctuation_to_remove_regex = f"[{''.join(punctuation_to_remove)}]"
+
+    if data_args.max_train_samples is not None:
+        raw_datasets["train"] = raw_datasets["train"].take(data_args.max_train_samples)
+
+    if data_args.max_eval_samples is not None:
+        raw_datasets["eval"] = raw_datasets["eval"].take(data_args.max_eval_samples)
+
+    def prepare_dataset(batch):
+        # process audio
+        sample = batch[audio_column_name]
+        inputs = feature_extractor(sample["array"], sampling_rate=sample["sampling_rate"])
+        # process audio length
+        batch[model_input_name] = inputs.get(model_input_name)[0]
+        batch["input_length"] = len(sample["array"])
+
+        # process targets (only char no rome)
+        input_str = batch[text_column_name].lower().split('（')[0] if do_lower_case else batch[text_column_name].split('（')[0]
+        
+        if do_remove_punctuation:
+            input_str = re.sub(punctuation_to_remove_regex, " ", input_str).strip()
+            input_str = re.sub("\s\s+", " ", input_str)
+        batch["labels"] = tokenizer(input_str).input_ids
+        return batch
+
+    with training_args.main_process_first(desc="dataset map pre-processing"):
+        vectorized_datasets = raw_datasets.map(
+            prepare_dataset,
+            remove_columns=raw_datasets_features,
+        ).with_format("torch")
+
+        if training_args.do_train:
+            vectorized_datasets["train"] = vectorized_datasets["train"].shuffle(
+                buffer_size=data_args.shuffle_buffer_size,
+                seed=training_args.seed,
+            )
+
+    # filter training data that is shorter than min_input_length or longer than
+    # max_input_length
+    def is_audio_in_length_range(length):
+        return min_input_length < length < max_input_length
+
+    if training_args.do_train:
+        vectorized_datasets["train"] = vectorized_datasets["train"].filter(
+            is_audio_in_length_range,
+            input_columns=["input_length"],
+        )
+
+    # 8. Load Metric
+    cer_metric = evaluate.load("cer")
+    wer_metric = evaluate.load("wer")
+    do_normalize_eval = data_args.do_normalize_eval
+
+    def compute_metrics(pred):
+        pred_ids = pred.predictions
+
+        pred.label_ids[pred.label_ids == -100] = tokenizer.pad_token_id
+
+        pred_str = tokenizer.batch_decode(pred_ids, skip_special_tokens=True, normalize=do_normalize_eval)
+        # we do not want to group tokens when computing the metrics
+        label_str = tokenizer.batch_decode(pred.label_ids, skip_special_tokens=True, normalize=do_normalize_eval)
+        cer = 100 * cer_metric.compute(predictions=pred_str, references=label_str)
+        pred_str = [ryNormText.separ_char_word(x) for x in pred_str]
+        label_str = [ryNormText.separ_char_word(x) for x in label_str]
+        p = open('pred.txt','w')
+        p.write(f'{pred_str=}')
+        p.close()
+        p = open('label.txt','w')
+        p.write(f'{label_str=}')
+        p.close()
+        wer = 100 * wer_metric.compute(predictions=pred_str, references=label_str)
+        return {"wer": wer, "cer": cer}
+
+    # 9. Create a single speech processor
+    if is_main_process(training_args.local_rank):
+        # save feature extractor, tokenizer and config
+        feature_extractor.save_pretrained(training_args.output_dir)
+        tokenizer.save_pretrained(training_args.output_dir)
+        config.save_pretrained(training_args.output_dir)
+
+    processor = AutoProcessor.from_pretrained(training_args.output_dir)
+
+    # 10. Define data collator
+    data_collator = DataCollatorSpeechSeq2SeqWithPadding(
+        processor=processor,
+        decoder_start_token_id=model.config.decoder_start_token_id,
+    )
+
+    # 11. Configure Trainer
+    # Trainer callback to reinitialise and reshuffle the streamable datasets at the beginning of each epoch
+    class ShuffleCallback(TrainerCallback):
+        def on_epoch_begin(self, args, state, control, train_dataloader, **kwargs):
+            if isinstance(train_dataloader.dataset, IterableDatasetShard):
+                pass  # set_epoch() is handled by the Trainer
+            elif isinstance(train_dataloader.dataset, IterableDataset):
+                train_dataloader.dataset.set_epoch(train_dataloader.dataset._epoch + 1)
+
+    # Initialize Trainer
+    trainer = Seq2SeqTrainer(
+        model=model,
+        args=training_args,
+        train_dataset=vectorized_datasets["train"] if training_args.do_train else None,
+        eval_dataset=vectorized_datasets["eval"] if training_args.do_eval else None,
+        tokenizer=feature_extractor,
+        data_collator=data_collator,
+        compute_metrics=compute_metrics if training_args.predict_with_generate else None,
+        callbacks=[ShuffleCallback()],
+    )
+
+    # 12. Training
+    if training_args.do_train:
+        checkpoint = None
+        if training_args.resume_from_checkpoint is not None:
+            checkpoint = training_args.resume_from_checkpoint
+        elif last_checkpoint is not None:
+            checkpoint = last_checkpoint
+        train_result = trainer.train(resume_from_checkpoint=checkpoint)
+        trainer.save_model()  # Saves the feature extractor too for easy upload
+
+        metrics = train_result.metrics
+        if data_args.max_train_samples:
+            metrics["train_samples"] = data_args.max_train_samples
+        trainer.log_metrics("train", metrics)
+        trainer.save_metrics("train", metrics)
+        trainer.save_state()
+
+    # 13. Evaluation
+    results = {}
+    if training_args.do_eval:
+        logger.info("*** Evaluate ***")
+        metrics = trainer.evaluate(
+
+            metric_key_prefix="eval",
+            max_length=training_args.generation_max_length,
+            num_beams=training_args.generation_num_beams,
+        )
+        if data_args.max_eval_samples:
+            metrics["eval_samples"] = data_args.max_eval_samples
+
+        trainer.log_metrics("eval", metrics)
+        trainer.save_metrics("eval", metrics)
+
+    # 14. Write Training Stats
+    kwargs = {
+        "finetuned_from": model_args.model_name_or_path,
+        "tasks": "automatic-speech-recognition",
+        "tags": "whisper-event",
+    }
+    if data_args.dataset_name is not None:
+        kwargs["dataset_tags"] = data_args.dataset_name
+        if data_args.dataset_config_name is not None:
+            kwargs["dataset"] = f"{data_args.dataset_name} {data_args.dataset_config_name}"
+        else:
+            kwargs["dataset"] = data_args.dataset_name
+        if "common_voice" in data_args.dataset_name:
+            kwargs["language"] = data_args.dataset_config_name
+        if model_args.model_index_name is not None:
+            kwargs["model_name"] = model_args.model_index_name
+
+    if training_args.push_to_hub:
+        trainer.push_to_hub(**kwargs)
+    else:
+        trainer.create_model_card(**kwargs)
+
+    return results
+
+
+if __name__ == "__main__":
+    main()

runs/May27_19-41-29_DESKTOP-EVSD799/1685187714.55857/events.out.tfevents.1685187714.DESKTOP-EVSD799.25803.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5500d1e2c8d5578bb18571a8d4d6cc288213b57d4c9b90efcdd866351cf4d973
+size 5980

runs/May27_19-41-29_DESKTOP-EVSD799/events.out.tfevents.1685187714.DESKTOP-EVSD799.25803.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff47ac59dcaf1ee4861f352b3a9fb2a99f19eabab7a0ac5819b82a0ea506daed
+size 38140

runs/May27_19-41-29_DESKTOP-EVSD799/events.out.tfevents.1685191671.DESKTOP-EVSD799.25803.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83d1f7e287620ce4d434d4b5130b733d03a4bc3fb0402bef362ca2334752b572
+size 405

ryNormText.py

@@ -0,0 +1,81 @@
+import string, re, opencc
+
+
+全型2半型= str.maketrans(
+    '　０１２３４５６７８９'
+    'ａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ'
+    'ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺ'
+    '！゛＃＄％＆（）＊＋、ー。／：；〈＝〉？＠［］＾＿‘｛｜｝～',
+    ' 0123456789'
+    'abcdefghijklmnopqrstuvwxyz'
+    'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
+    '!"#$%&()*+,-./:;<=>?@[]^_`{|}~'
+    )
+
+def 把怪字修進unicode(xStr):
+    xStr= re.sub('\uf5c3','𪜶', xStr)
+    return xStr
+
+def ryNormText(s):
+    """
+    <<<None>>> ==> 刪除
+    標點 ==> 空白
+    連續空白 ==> 1個空白
+    簡繁
+    """
+
+    punc1= string.punctuation  # 這是為英文
+    punc1
+    punc2= '。，﹐、！？：：；『』「」…,\n' # 這是為中文，尚未完備！！
+
+    punc= f"[{punc1}{punc2}]"  ## 這是 regular expression 的 pattern
+
+    ## <<<None>>> ==> 刪除
+    s= re.sub('<<<None>>>','',s)
+
+    # 標點 ==> 空白
+    s= re.sub(punc,' ',s)
+    
+
+    # 連續空白 ==> 1個空白
+    s= re.sub('[ ]+',' ',s)
+    
+    
+    # 空白 ==> 刪除
+    s= re.sub(' ','',s)
+    
+    s= 把怪字修進unicode(s)
+
+    # 簡繁
+    s= opencc.OpenCC('s2tw').convert(s)
+    
+
+    return s
+
+import unicodedata
+import re
+
+
+def separ_char_word(inputString= '我是呂仁園 Renyuan Lyu'):
+
+    inputString= 把怪字修進unicode(inputString)
+
+    y= ''
+    for x in inputString:
+        y += x
+        try:
+            un= unicodedata.name(x)
+            if un.startswith('CJK'):
+                y += ' '
+            else:
+                pass
+        except Exception as ex:
+            y = ' '+y+' '
+            print(f'ryErr:(def 中英分開:){ex= }\t【{x= }】\t{inputString= }')
+
+    y= re.sub('[ ]+',' ', y) #連續空白只保留1個空白
+    return y
+
+#q= 中英分開('大家好 da jia hao 我是呂仁園 I am Renyuan Lyu')
+
+#print(q)

special_tokens_map.json

@@ -0,0 +1,133 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<|startoftranscript|>",
+    "<|en|>",
+    "<|zh|>",
+    "<|de|>",
+    "<|es|>",
+    "<|ru|>",
+    "<|ko|>",
+    "<|fr|>",
+    "<|ja|>",
+    "<|pt|>",
+    "<|tr|>",
+    "<|pl|>",
+    "<|ca|>",
+    "<|nl|>",
+    "<|ar|>",
+    "<|sv|>",
+    "<|it|>",
+    "<|id|>",
+    "<|hi|>",
+    "<|fi|>",
+    "<|vi|>",
+    "<|he|>",
+    "<|uk|>",
+    "<|el|>",
+    "<|ms|>",
+    "<|cs|>",
+    "<|ro|>",
+    "<|da|>",
+    "<|hu|>",
+    "<|ta|>",
+    "<|no|>",
+    "<|th|>",
+    "<|ur|>",
+    "<|hr|>",
+    "<|bg|>",
+    "<|lt|>",
+    "<|la|>",
+    "<|mi|>",
+    "<|ml|>",
+    "<|cy|>",
+    "<|sk|>",
+    "<|te|>",
+    "<|fa|>",
+    "<|lv|>",
+    "<|bn|>",
+    "<|sr|>",
+    "<|az|>",
+    "<|sl|>",
+    "<|kn|>",
+    "<|et|>",
+    "<|mk|>",
+    "<|br|>",
+    "<|eu|>",
+    "<|is|>",
+    "<|hy|>",
+    "<|ne|>",
+    "<|mn|>",
+    "<|bs|>",
+    "<|kk|>",
+    "<|sq|>",
+    "<|sw|>",
+    "<|gl|>",
+    "<|mr|>",
+    "<|pa|>",
+    "<|si|>",
+    "<|km|>",
+    "<|sn|>",
+    "<|yo|>",
+    "<|so|>",
+    "<|af|>",
+    "<|oc|>",
+    "<|ka|>",
+    "<|be|>",
+    "<|tg|>",
+    "<|sd|>",
+    "<|gu|>",
+    "<|am|>",
+    "<|yi|>",
+    "<|lo|>",
+    "<|uz|>",
+    "<|fo|>",
+    "<|ht|>",
+    "<|ps|>",
+    "<|tk|>",
+    "<|nn|>",
+    "<|mt|>",
+    "<|sa|>",
+    "<|lb|>",
+    "<|my|>",
+    "<|bo|>",
+    "<|tl|>",
+    "<|mg|>",
+    "<|as|>",
+    "<|tt|>",
+    "<|haw|>",
+    "<|ln|>",
+    "<|ha|>",
+    "<|ba|>",
+    "<|jw|>",
+    "<|su|>",
+    "<|translate|>",
+    "<|transcribe|>",
+    "<|startoflm|>",
+    "<|startofprev|>",
+    "<|nocaptions|>",
+    "<|notimestamps|>"
+  ],
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|endoftext|>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,35 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "errors": "replace",
+  "model_max_length": 1024,
+  "pad_token": null,
+  "processor_class": "WhisperProcessor",
+  "return_attention_mask": false,
+  "special_tokens_map_file": null,
+  "tokenizer_class": "WhisperTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 6.05,
+    "train_loss": 0.3284327008752851,
+    "train_runtime": 3725.671,
+    "train_samples_per_second": 2.684,
+    "train_steps_per_second": 1.342
+}

trainer_state.json

@@ -0,0 +1,1275 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 6.0532,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2000000000000006e-07,
+      "loss": 2.3482,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.200000000000001e-07,
+      "loss": 2.3433,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.42e-06,
+      "loss": 1.761,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 1.5988,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.4245,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9e-06,
+      "loss": 1.5665,
+      "step": 150
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.4588,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 1.2609,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.4e-06,
+      "loss": 1.2301,
+      "step": 225
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.4295,
+      "step": 250
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.415,
+      "step": 275
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9e-06,
+      "loss": 1.1804,
+      "step": 300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.3484,
+      "step": 325
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.9e-06,
+      "loss": 1.3593,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 7.4e-06,
+      "loss": 1.179,
+      "step": 375
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.9e-06,
+      "loss": 1.3824,
+      "step": 400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.1744,
+      "step": 425
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 1.2079,
+      "step": 450
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.4e-06,
+      "loss": 1.2476,
+      "step": 475
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.9e-06,
+      "loss": 1.3404,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.955555555555556e-06,
+      "loss": 1.0947,
+      "step": 525
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.9e-06,
+      "loss": 1.3787,
+      "step": 550
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.844444444444446e-06,
+      "loss": 1.0917,
+      "step": 575
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.78888888888889e-06,
+      "loss": 1.1884,
+      "step": 600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.733333333333334e-06,
+      "loss": 1.2814,
+      "step": 625
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.677777777777778e-06,
+      "loss": 1.1693,
+      "step": 650
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.622222222222222e-06,
+      "loss": 1.3852,
+      "step": 675
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.566666666666668e-06,
+      "loss": 1.0322,
+      "step": 700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.511111111111112e-06,
+      "loss": 1.1455,
+      "step": 725
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.455555555555557e-06,
+      "loss": 0.9578,
+      "step": 750
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.4e-06,
+      "loss": 1.2149,
+      "step": 775
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.344444444444446e-06,
+      "loss": 0.9243,
+      "step": 800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.28888888888889e-06,
+      "loss": 0.7373,
+      "step": 825
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.233333333333334e-06,
+      "loss": 0.7298,
+      "step": 850
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.17777777777778e-06,
+      "loss": 0.7556,
+      "step": 875
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.122222222222223e-06,
+      "loss": 0.6975,
+      "step": 900
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 0.7299,
+      "step": 925
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.011111111111111e-06,
+      "loss": 0.6555,
+      "step": 950
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.955555555555555e-06,
+      "loss": 0.6755,
+      "step": 975
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 0.5832,
+      "step": 1000
+    },
+    {
+      "epoch": 1.04,
+      "eval_cer": 56.474494415937215,
+      "eval_loss": 1.0634080171585083,
+      "eval_runtime": 219.1746,
+      "eval_samples_per_second": 4.499,
+      "eval_steps_per_second": 2.249,
+      "eval_wer": 56.30534351145038,
+      "step": 1000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 8.844444444444445e-06,
+      "loss": 0.5757,
+      "step": 1025
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 8.788888888888891e-06,
+      "loss": 0.6065,
+      "step": 1050
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 8.733333333333333e-06,
+      "loss": 0.5487,
+      "step": 1075
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 8.677777777777779e-06,
+      "loss": 0.6339,
+      "step": 1100
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 8.622222222222223e-06,
+      "loss": 0.5477,
+      "step": 1125
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 8.566666666666667e-06,
+      "loss": 0.597,
+      "step": 1150
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 8.511111111111113e-06,
+      "loss": 0.5394,
+      "step": 1175
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 8.455555555555555e-06,
+      "loss": 0.52,
+      "step": 1200
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.6005,
+      "step": 1225
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 8.344444444444445e-06,
+      "loss": 0.4988,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 8.288888888888889e-06,
+      "loss": 0.5381,
+      "step": 1275
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 8.233333333333335e-06,
+      "loss": 0.5005,
+      "step": 1300
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 8.177777777777779e-06,
+      "loss": 0.3982,
+      "step": 1325
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 8.122222222222223e-06,
+      "loss": 0.5537,
+      "step": 1350
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 8.066666666666667e-06,
+      "loss": 0.3985,
+      "step": 1375
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 8.011111111111113e-06,
+      "loss": 0.3653,
+      "step": 1400
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.955555555555557e-06,
+      "loss": 0.4001,
+      "step": 1425
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.9e-06,
+      "loss": 0.4278,
+      "step": 1450
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.844444444444446e-06,
+      "loss": 0.3982,
+      "step": 1475
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.788888888888889e-06,
+      "loss": 0.3967,
+      "step": 1500
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 0.4366,
+      "step": 1525
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 7.677777777777778e-06,
+      "loss": 0.4665,
+      "step": 1550
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 7.622222222222223e-06,
+      "loss": 0.4292,
+      "step": 1575
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.566666666666667e-06,
+      "loss": 0.2518,
+      "step": 1600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.511111111111111e-06,
+      "loss": 0.1433,
+      "step": 1625
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.455555555555556e-06,
+      "loss": 0.1932,
+      "step": 1650
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.4e-06,
+      "loss": 0.1875,
+      "step": 1675
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.344444444444445e-06,
+      "loss": 0.2236,
+      "step": 1700
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.28888888888889e-06,
+      "loss": 0.2445,
+      "step": 1725
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.233333333333334e-06,
+      "loss": 0.1751,
+      "step": 1750
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.177777777777778e-06,
+      "loss": 0.1976,
+      "step": 1775
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.122222222222222e-06,
+      "loss": 0.1307,
+      "step": 1800
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 7.066666666666667e-06,
+      "loss": 0.1849,
+      "step": 1825
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 7.011111111111112e-06,
+      "loss": 0.1233,
+      "step": 1850
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.955555555555557e-06,
+      "loss": 0.136,
+      "step": 1875
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.9e-06,
+      "loss": 0.1014,
+      "step": 1900
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.844444444444445e-06,
+      "loss": 0.1366,
+      "step": 1925
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.788888888888889e-06,
+      "loss": 0.2011,
+      "step": 1950
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.733333333333334e-06,
+      "loss": 0.0962,
+      "step": 1975
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.677777777777779e-06,
+      "loss": 0.1467,
+      "step": 2000
+    },
+    {
+      "epoch": 2.08,
+      "eval_cer": 51.01116812556595,
+      "eval_loss": 1.040711760520935,
+      "eval_runtime": 219.0004,
+      "eval_samples_per_second": 4.502,
+      "eval_steps_per_second": 2.251,
+      "eval_wer": 50.961832061068705,
+      "step": 2000
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.6222222222222236e-06,
+      "loss": 0.1376,
+      "step": 2025
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.566666666666667e-06,
+      "loss": 0.1322,
+      "step": 2050
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.511111111111112e-06,
+      "loss": 0.0889,
+      "step": 2075
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.455555555555556e-06,
+      "loss": 0.0725,
+      "step": 2100
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.084,
+      "step": 2125
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 6.3444444444444454e-06,
+      "loss": 0.078,
+      "step": 2150
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 6.28888888888889e-06,
+      "loss": 0.0666,
+      "step": 2175
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 6.2333333333333335e-06,
+      "loss": 0.1335,
+      "step": 2200
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 6.177777777777778e-06,
+      "loss": 0.0919,
+      "step": 2225
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 6.1222222222222224e-06,
+      "loss": 0.0966,
+      "step": 2250
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 6.066666666666667e-06,
+      "loss": 0.0947,
+      "step": 2275
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 6.011111111111112e-06,
+      "loss": 0.0731,
+      "step": 2300
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.955555555555555e-06,
+      "loss": 0.1585,
+      "step": 2325
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.9e-06,
+      "loss": 0.1475,
+      "step": 2350
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5.844444444444445e-06,
+      "loss": 0.0635,
+      "step": 2375
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.788888888888889e-06,
+      "loss": 0.0489,
+      "step": 2400
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.733333333333334e-06,
+      "loss": 0.0325,
+      "step": 2425
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5.677777777777779e-06,
+      "loss": 0.0661,
+      "step": 2450
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5.622222222222222e-06,
+      "loss": 0.0433,
+      "step": 2475
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5.566666666666667e-06,
+      "loss": 0.0215,
+      "step": 2500
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5.511111111111112e-06,
+      "loss": 0.0477,
+      "step": 2525
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5.455555555555556e-06,
+      "loss": 0.0315,
+      "step": 2550
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.0798,
+      "step": 2575
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5.344444444444446e-06,
+      "loss": 0.0298,
+      "step": 2600
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5.288888888888889e-06,
+      "loss": 0.033,
+      "step": 2625
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5.233333333333334e-06,
+      "loss": 0.0151,
+      "step": 2650
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5.177777777777779e-06,
+      "loss": 0.0465,
+      "step": 2675
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5.122222222222223e-06,
+      "loss": 0.0138,
+      "step": 2700
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 0.0293,
+      "step": 2725
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5.011111111111111e-06,
+      "loss": 0.0296,
+      "step": 2750
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 4.9555555555555565e-06,
+      "loss": 0.0336,
+      "step": 2775
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0377,
+      "step": 2800
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 4.8444444444444446e-06,
+      "loss": 0.0345,
+      "step": 2825
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 4.7888888888888894e-06,
+      "loss": 0.0129,
+      "step": 2850
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 4.7333333333333335e-06,
+      "loss": 0.0141,
+      "step": 2875
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 4.677777777777778e-06,
+      "loss": 0.0165,
+      "step": 2900
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 4.622222222222222e-06,
+      "loss": 0.0318,
+      "step": 2925
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 4.566666666666667e-06,
+      "loss": 0.0161,
+      "step": 2950
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 4.511111111111111e-06,
+      "loss": 0.028,
+      "step": 2975
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 4.455555555555555e-06,
+      "loss": 0.016,
+      "step": 3000
+    },
+    {
+      "epoch": 3.13,
+      "eval_cer": 46.5137337760338,
+      "eval_loss": 1.0225664377212524,
+      "eval_runtime": 217.5747,
+      "eval_samples_per_second": 4.532,
+      "eval_steps_per_second": 2.266,
+      "eval_wer": 46.44274809160305,
+      "step": 3000
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0509,
+      "step": 3025
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.344444444444445e-06,
+      "loss": 0.0254,
+      "step": 3050
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.288888888888889e-06,
+      "loss": 0.0075,
+      "step": 3075
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 4.233333333333334e-06,
+      "loss": 0.0303,
+      "step": 3100
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 4.177777777777778e-06,
+      "loss": 0.0127,
+      "step": 3125
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 4.122222222222222e-06,
+      "loss": 0.0184,
+      "step": 3150
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.066666666666667e-06,
+      "loss": 0.0311,
+      "step": 3175
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.011111111111111e-06,
+      "loss": 0.0012,
+      "step": 3200
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.955555555555556e-06,
+      "loss": 0.0106,
+      "step": 3225
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.0085,
+      "step": 3250
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.844444444444445e-06,
+      "loss": 0.0016,
+      "step": 3275
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.7888888888888893e-06,
+      "loss": 0.0083,
+      "step": 3300
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 0.0099,
+      "step": 3325
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.6777777777777778e-06,
+      "loss": 0.0017,
+      "step": 3350
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.6222222222222226e-06,
+      "loss": 0.0097,
+      "step": 3375
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.566666666666667e-06,
+      "loss": 0.0083,
+      "step": 3400
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.511111111111111e-06,
+      "loss": 0.0182,
+      "step": 3425
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.455555555555556e-06,
+      "loss": 0.0041,
+      "step": 3450
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.0321,
+      "step": 3475
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.3444444444444445e-06,
+      "loss": 0.0009,
+      "step": 3500
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.2888888888888894e-06,
+      "loss": 0.0008,
+      "step": 3525
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.2333333333333334e-06,
+      "loss": 0.0016,
+      "step": 3550
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.177777777777778e-06,
+      "loss": 0.0368,
+      "step": 3575
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.1222222222222228e-06,
+      "loss": 0.006,
+      "step": 3600
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.066666666666667e-06,
+      "loss": 0.0023,
+      "step": 3625
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.0111111111111113e-06,
+      "loss": 0.001,
+      "step": 3650
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 2.955555555555556e-06,
+      "loss": 0.0012,
+      "step": 3675
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0009,
+      "step": 3700
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.8444444444444446e-06,
+      "loss": 0.0009,
+      "step": 3725
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.788888888888889e-06,
+      "loss": 0.0005,
+      "step": 3750
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.7333333333333336e-06,
+      "loss": 0.0095,
+      "step": 3775
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.677777777777778e-06,
+      "loss": 0.0072,
+      "step": 3800
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.6222222222222225e-06,
+      "loss": 0.0203,
+      "step": 3825
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.566666666666667e-06,
+      "loss": 0.001,
+      "step": 3850
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.5111111111111114e-06,
+      "loss": 0.0004,
+      "step": 3875
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.455555555555556e-06,
+      "loss": 0.0224,
+      "step": 3900
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.0012,
+      "step": 3925
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.3444444444444448e-06,
+      "loss": 0.0002,
+      "step": 3950
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.2888888888888892e-06,
+      "loss": 0.0003,
+      "step": 3975
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.2333333333333333e-06,
+      "loss": 0.0001,
+      "step": 4000
+    },
+    {
+      "epoch": 5.01,
+      "eval_cer": 45.60821008149713,
+      "eval_loss": 0.9974298477172852,
+      "eval_runtime": 214.4422,
+      "eval_samples_per_second": 4.598,
+      "eval_steps_per_second": 2.299,
+      "eval_wer": 45.465648854961835,
+      "step": 4000
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.1777777777777777e-06,
+      "loss": 0.0002,
+      "step": 4025
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.1222222222222226e-06,
+      "loss": 0.0003,
+      "step": 4050
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.0666666666666666e-06,
+      "loss": 0.0011,
+      "step": 4075
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.011111111111111e-06,
+      "loss": 0.0002,
+      "step": 4100
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.955555555555556e-06,
+      "loss": 0.0005,
+      "step": 4125
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.0003,
+      "step": 4150
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1.8444444444444445e-06,
+      "loss": 0.0001,
+      "step": 4175
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1.788888888888889e-06,
+      "loss": 0.0002,
+      "step": 4200
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1.7333333333333336e-06,
+      "loss": 0.0018,
+      "step": 4225
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1.6777777777777779e-06,
+      "loss": 0.0002,
+      "step": 4250
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.6222222222222223e-06,
+      "loss": 0.0001,
+      "step": 4275
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.566666666666667e-06,
+      "loss": 0.0037,
+      "step": 4300
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1.5111111111111112e-06,
+      "loss": 0.0002,
+      "step": 4325
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1.4555555555555557e-06,
+      "loss": 0.0001,
+      "step": 4350
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.0001,
+      "step": 4375
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1.3444444444444446e-06,
+      "loss": 0.0004,
+      "step": 4400
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1.288888888888889e-06,
+      "loss": 0.0001,
+      "step": 4425
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1.2333333333333335e-06,
+      "loss": 0.0001,
+      "step": 4450
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.1777777777777778e-06,
+      "loss": 0.0001,
+      "step": 4475
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.1222222222222222e-06,
+      "loss": 0.0005,
+      "step": 4500
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 0.0014,
+      "step": 4525
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1.0111111111111111e-06,
+      "loss": 0.0001,
+      "step": 4550
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 9.555555555555556e-07,
+      "loss": 0.0022,
+      "step": 4575
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.0001,
+      "step": 4600
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 8.444444444444445e-07,
+      "loss": 0.0003,
+      "step": 4625
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.888888888888889e-07,
+      "loss": 0.0001,
+      "step": 4650
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.333333333333334e-07,
+      "loss": 0.0003,
+      "step": 4675
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 6.777777777777779e-07,
+      "loss": 0.0001,
+      "step": 4700
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 6.222222222222223e-07,
+      "loss": 0.0001,
+      "step": 4725
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 5.666666666666667e-07,
+      "loss": 0.0001,
+      "step": 4750
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 5.111111111111112e-07,
+      "loss": 0.0001,
+      "step": 4775
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.5555555555555563e-07,
+      "loss": 0.0001,
+      "step": 4800
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.0001,
+      "step": 4825
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 3.444444444444445e-07,
+      "loss": 0.0001,
+      "step": 4850
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.888888888888889e-07,
+      "loss": 0.0001,
+      "step": 4875
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.3333333333333336e-07,
+      "loss": 0.0001,
+      "step": 4900
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 1.777777777777778e-07,
+      "loss": 0.0001,
+      "step": 4925
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 1.2222222222222225e-07,
+      "loss": 0.0001,
+      "step": 4950
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 6.666666666666668e-08,
+      "loss": 0.0001,
+      "step": 4975
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 1.1111111111111112e-08,
+      "loss": 0.0001,
+      "step": 5000
+    },
+    {
+      "epoch": 6.05,
+      "eval_cer": 45.36673709628735,
+      "eval_loss": 0.994395911693573,
+      "eval_runtime": 216.4096,
+      "eval_samples_per_second": 4.556,
+      "eval_steps_per_second": 2.278,
+      "eval_wer": 45.2824427480916,
+      "step": 5000
+    },
+    {
+      "epoch": 6.05,
+      "step": 5000,
+      "total_flos": 1.019992544575488e+19,
+      "train_loss": 0.3284327008752851,
+      "train_runtime": 3725.671,
+      "train_samples_per_second": 2.684,
+      "train_steps_per_second": 1.342
+    }
+  ],
+  "max_steps": 5000,
+  "num_train_epochs": 9223372036854775807,
+  "total_flos": 1.019992544575488e+19,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:597458d84bfea0d6498124c77965aabd1bbaac3387cdb9ee39c08874f88302dd
+size 3643