Fix eval script

README.md

@@ -82,37 +82,38 @@ from tqdm.auto import tqdm
 from datasets import load_metric
 from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
 
-# Download the raw data instead of using HF datasets to save space 
-data_url = "https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-6.1-2020-12-11/sah.tar.gz"
+# Download the raw data instead of using HF datasets to save disk space 
+data_url = "https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-6.1-2020-12-11/lt.tar.gz"
 filestream = urllib.request.urlopen(data_url)
 data_file = tarfile.open(fileobj=filestream, mode="r|gz")
 data_file.extractall()
 
 wer = load_metric("wer")
 
-processor = Wav2Vec2Processor.from_pretrained("anton-l/wav2vec2-large-xlsr-53-sakha")
-model = Wav2Vec2ForCTC.from_pretrained("anton-l/wav2vec2-large-xlsr-53-sakha")
+processor = Wav2Vec2Processor.from_pretrained("anton-l/wav2vec2-large-xlsr-53-lithuanian")
+model = Wav2Vec2ForCTC.from_pretrained("anton-l/wav2vec2-large-xlsr-53-lithuanian")
 model.to("cuda")
 
-cv_test = pd.read_csv("cv-corpus-6.1-2020-12-11/sah/test.tsv", sep='\\t')
-clips_path = "cv-corpus-6.1-2020-12-11/sah/clips/"
+cv_test = pd.read_csv("cv-corpus-6.1-2020-12-11/lt/test.tsv", sep='\t')
+clips_path = "cv-corpus-6.1-2020-12-11/lt/clips/"
 
 def clean_sentence(sent):
     sent = sent.lower()
+    # normalize apostrophes
+    sent = sent.replace("’", "'")
     # replace non-alpha characters with space
-    sent = "".join(ch if ch.isalpha() else " " for ch in sent)
+    sent = "".join(ch if ch.isalpha() or ch == "'" else " " for ch in sent)
     # remove repeated spaces
     sent = " ".join(sent.split())
     return sent
 
-resampler = torchaudio.transforms.Resample(48_000, 16_000)
-
 targets = []
 preds = []
 
 for i, row in tqdm(cv_test.iterrows(), total=cv_test.shape[0]):
     row["sentence"] = clean_sentence(row["sentence"])
     speech_array, sampling_rate = torchaudio.load(clips_path + row["path"])
+    resampler = torchaudio.transforms.Resample(sampling_rate, 16_000)
     row["speech"] = resampler(speech_array).squeeze().numpy()
 
     inputs = processor(row["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)