Update README.md
Browse files
README.md
CHANGED
|
@@ -27,7 +27,9 @@ model-index:
|
|
| 27 |
# Wav2Vec2-Large-XLSR-Upper-Sorbian
|
| 28 |
|
| 29 |
Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53)
|
| 30 |
-
on the [Upper Sorbian Common Voice dataset](https://huggingface.co/datasets/common_voice)
|
|
|
|
|
|
|
| 31 |
When using this model, make sure that your speech input is sampled at 16kHz.
|
| 32 |
|
| 33 |
## Usage
|
|
@@ -75,7 +77,7 @@ wer = load_metric("wer")
|
|
| 75 |
processor = Wav2Vec2Processor.from_pretrained("jimregan/wav2vec2-large-xlsr-upper-sorbian-mixed")
|
| 76 |
model = Wav2Vec2ForCTC.from_pretrained("jimregan/wav2vec2-large-xlsr-upper-sorbian-mixed")
|
| 77 |
model.to("cuda")
|
| 78 |
-
chars_to_ignore_regex = '[
|
| 79 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 80 |
# Preprocessing the datasets.
|
| 81 |
# We need to read the audio files as arrays
|
|
|
|
| 27 |
# Wav2Vec2-Large-XLSR-Upper-Sorbian
|
| 28 |
|
| 29 |
Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53)
|
| 30 |
+
on the [Upper Sorbian Common Voice dataset](https://huggingface.co/datasets/common_voice), with an
|
| 31 |
+
extra 28 minutes of audio from an online [Sorbian course](https://sprachkurs.sorbischlernen.de/).
|
| 32 |
+
|
| 33 |
When using this model, make sure that your speech input is sampled at 16kHz.
|
| 34 |
|
| 35 |
## Usage
|
|
|
|
| 77 |
processor = Wav2Vec2Processor.from_pretrained("jimregan/wav2vec2-large-xlsr-upper-sorbian-mixed")
|
| 78 |
model = Wav2Vec2ForCTC.from_pretrained("jimregan/wav2vec2-large-xlsr-upper-sorbian-mixed")
|
| 79 |
model.to("cuda")
|
| 80 |
+
chars_to_ignore_regex = '[\\\\\\\\,\\\\\\\\?\\\\\\\\.\\\\\\\\!\\\\\\\\-\\\\\\\\;\\\\\\\\:\\\\\\\\"\\\\\\\\“\\\\\\\\%\\\\\\\\‘\\\\\\\\”\\\\\\\\�„«»–]'
|
| 81 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 82 |
# Preprocessing the datasets.
|
| 83 |
# We need to read the audio files as arrays
|