Upload lm-boosted decoder new

alphabet.json

@@ -0,0 +1 @@
+{"labels": [" ", "c", "h", "k", "o", "s", "t", "\u2047", "", "<s>", "</s>"], "is_bpe": false}

language_model/attrs.json

@@ -0,0 +1 @@
+{"alpha": 0.5, "beta": 1.5, "unk_score_offset": -10.0, "score_boundary": true}

language_model/unigrams.txt

@@ -0,0 +1,27 @@
+</s>
+<s>
+c
+c.
+ch
+ck
+ck.
+cs
+cs.
+h
+h.
+k
+k.
+kh
+kh.
+o
+o.
+s
+s.
+sh
+sh.
+sk
+skh
+ss
+st
+t
+t.

preprocessor_config.json

@@ -4,7 +4,7 @@
   "feature_size": 1,
   "padding_side": "right",
   "padding_value": 0.0,
-  "processor_class": "Wav2Vec2Processor",
+  "processor_class": "Wav2Vec2ProcessorWithLM",
   "return_attention_mask": false,
   "sampling_rate": 16000
 }

special_tokens_map.json

@@ -21,6 +21,20 @@
       "rstrip": false,
       "single_word": false
     },
+    {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false
+    },
+    {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false
+    },
     {
       "content": "</s>",
       "lstrip": false,

tokenizer_config.json

@@ -2,9 +2,9 @@
   "bos_token": "<s>",
   "do_lower_case": false,
   "eos_token": "</s>",
-  "name_or_path": "ai-light-dance_drums_ft_pretrain_wav2vec2-base-new-v2",
+  "name_or_path": "gary109/ai-light-dance_drums_ft_pretrain_wav2vec2-base-new-v2",
   "pad_token": "[PAD]",
-  "processor_class": "Wav2Vec2Processor",
+  "processor_class": "Wav2Vec2ProcessorWithLM",
   "replace_word_delimiter_char": " ",
   "special_tokens_map_file": null,
   "tokenizer_class": "Wav2Vec2CTCTokenizer",