Update README.md

README.md

@@ -57,8 +57,8 @@ def load_dataset_sundanese():
     
     dfs = []
     
-    dfs.append(pd.read_csv(filenames[0], sep='\\\\\\\\t\\\\\\\\t', names=["path", "sentence"]))
-    dfs.append(pd.read_csv(filenames[1], sep='\\\\\\\\t\\\\\\\\t', names=["path", "sentence"]))
+    dfs.append(pd.read_csv(filenames[0], sep='\\\\\\\\\\\\\\\\t\\\\\\\\\\\\\\\\t', names=["path", "sentence"]))
+    dfs.append(pd.read_csv(filenames[1], sep='\\\\\\\\\\\\\\\\t\\\\\\\\\\\\\\\\t', names=["path", "sentence"]))
     
     for i, dir in enumerate(data_dirs):
         dfs[i]["path"] = dfs[i].apply(lambda row: str(data_dirs[i]) + "/" + row + ".wav", axis=1)
@@ -124,8 +124,8 @@ def load_dataset_sundanese():
     
     dfs = []
     
-    dfs.append(pd.read_csv(filenames[0], sep='\\\\\\\\t\\\\\\\\t', names=["path", "sentence"]))
-    dfs.append(pd.read_csv(filenames[1], sep='\\\\\\\\t\\\\\\\\t', names=["path", "sentence"]))
+    dfs.append(pd.read_csv(filenames[0], sep='\\\\\\\\\\\\\\\\t\\\\\\\\\\\\\\\\t', names=["path", "sentence"]))
+    dfs.append(pd.read_csv(filenames[1], sep='\\\\\\\\\\\\\\\\t\\\\\\\\\\\\\\\\t', names=["path", "sentence"]))
     
     for i, dir in enumerate(data_dirs):
         dfs[i]["path"] = dfs[i].apply(lambda row: str(data_dirs[i]) + "/" + row + ".wav", axis=1)
@@ -145,7 +145,7 @@ processor = Wav2Vec2Processor.from_pretrained("cahya/wav2vec2-large-xlsr-sundane
 model = Wav2Vec2ForCTC.from_pretrained("cahya/wav2vec2-large-xlsr-sundanese") 
 model.to("cuda")
 
-chars_to_ignore_regex = '[\\,\\?\\.\\!\\-\\;\\:\\"\\“\\%\\‘\\'\\”_\\�]'
+chars_to_ignore_regex = '[\\\\,\\\\?\\\\.\\\\!\\\\-\\\\;\\\\:\\\\"\\\\“\\\\%\\\\‘\\\\'\\\\”_\\\\�]'
 resampler = torchaudio.transforms.Resample(48_000, 16_000)
 
 # Preprocessing the datasets.
@@ -179,7 +179,6 @@ print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"],
 
 ## Training
 
-The Common Voice `train`, `validation`, and ... datasets were used for training as well as ... and ...  # TODO
-
+[OpenSLR High quality TTS data for Sundanese](https://openslr.org/44/) was used for training.
 The script used for training can be found [here](https://github.com/cahya-wirawan/indonesian-speech-recognition/blob/main/XLSR_Wav2Vec2_for_Indonesian_Evaluation-Sundanese.ipynb) 
 and to [evaluate it](https://github.com/cahya-wirawan/indonesian-speech-recognition/blob/main/XLSR_Wav2Vec2_for_Indonesian_Evaluation-Sundanese.ipynb)