versae commited on
Commit
e43c673
1 Parent(s): c9e5f00

Add bokmaal 5gram model

Browse files
.gitattributes CHANGED
@@ -31,3 +31,4 @@ wandb/run-20220523_115145-3dybzmyz/logs/debug-internal.log filter=lfs diff=lfs m
31
  wandb/run-20220616_112025-1526e2dp/logs/debug-internal.log filter=lfs diff=lfs merge=lfs -text
32
  wandb/run-20220622_135628-3qxrqni1/logs/debug-internal.log filter=lfs diff=lfs merge=lfs -text
33
  wandb/run-20220622_135628-3qxrqni1/files/output.log filter=lfs diff=lfs merge=lfs -text
 
 
31
  wandb/run-20220616_112025-1526e2dp/logs/debug-internal.log filter=lfs diff=lfs merge=lfs -text
32
  wandb/run-20220622_135628-3qxrqni1/logs/debug-internal.log filter=lfs diff=lfs merge=lfs -text
33
  wandb/run-20220622_135628-3qxrqni1/files/output.log filter=lfs diff=lfs merge=lfs -text
34
+ language_model/unigrams.txt filter=lfs diff=lfs merge=lfs -text
add_kenlm.py ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import argparse
2
+ from transformers import AutoProcessor
3
+ from transformers import Wav2Vec2ProcessorWithLM
4
+ from pyctcdecode import build_ctcdecoder
5
+
6
+
7
+ def main(args):
8
+ processor = AutoProcessor.from_pretrained(args.model_name_or_path)
9
+ vocab_dict = processor.tokenizer.get_vocab()
10
+ sorted_vocab_dict = {
11
+ k.lower(): v for k, v in sorted(vocab_dict.items(), key=lambda item: item[1])
12
+ }
13
+ decoder = build_ctcdecoder(
14
+ labels=list(sorted_vocab_dict.keys()),
15
+ kenlm_model_path=args.kenlm_model_path,
16
+ )
17
+ processor_with_lm = Wav2Vec2ProcessorWithLM(
18
+ feature_extractor=processor.feature_extractor,
19
+ tokenizer=processor.tokenizer,
20
+ decoder=decoder,
21
+ )
22
+ processor_with_lm.save_pretrained(args.model_name_or_path)
23
+ print(
24
+ f"Run: ~/bin/build_binary language_model/*.arpa language_model/5gram.bin -T $(pwd) && rm language_model/*.arpa")
25
+
26
+
27
+ def parse_args():
28
+ parser = argparse.ArgumentParser()
29
+ parser.add_argument('--model_name_or_path', default="./", help='Model name or path. Defaults to ./')
30
+ parser.add_argument('--kenlm_model_path', required=True, help='Path to KenLM arpa file.')
31
+ args = parser.parse_args()
32
+ return args
33
+
34
+
35
+ if __name__ == "__main__":
36
+ main(parse_args())
37
+
alphabet.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"labels": [" ", "a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z", "\u00e5", "\u00e6", "\u00f8", "\u2047", "", "<s>", "</s>"], "is_bpe": false}
language_model/5gram.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b41c24c63f2f0585bea83666369593f3b3e6d047f327a90f36ebca2c35ef0ff
3
+ size 4243671427
language_model/attrs.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"alpha": 0.5, "beta": 1.5, "unk_score_offset": -10.0, "score_boundary": true}
language_model/unigrams.txt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac3e71ca49838ca355df6fdcb8d89344a5a9bf9e1a76587cdf5df1367c19b9a9
3
+ size 16759269
preprocessor_config.json CHANGED
@@ -4,6 +4,7 @@
4
  "feature_size": 1,
5
  "padding_side": "right",
6
  "padding_value": 0,
 
7
  "return_attention_mask": true,
8
  "sampling_rate": 16000
9
  }
 
4
  "feature_size": 1,
5
  "padding_side": "right",
6
  "padding_value": 0,
7
+ "processor_class": "Wav2Vec2ProcessorWithLM",
8
  "return_attention_mask": true,
9
  "sampling_rate": 16000
10
  }
special_tokens_map.json CHANGED
@@ -1 +1 @@
1
- {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "additional_special_tokens": [{"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "additional_special_tokens": [{"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}
tokenizer_config.json CHANGED
@@ -1 +1 @@
1
- {"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "replace_word_delimiter_char": " ", "special_tokens_map_file": null, "name_or_path": "./", "tokenizer_class": "Wav2Vec2CTCTokenizer"}
 
1
+ {"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "replace_word_delimiter_char": " ", "special_tokens_map_file": null, "name_or_path": ".", "tokenizer_class": "Wav2Vec2CTCTokenizer", "processor_class": "Wav2Vec2ProcessorWithLM"}