Merge branch 'main' of https://huggingface.co/alexcleu/wav2vec2-large-xlsr-polish into main

README.md

@@ -9,7 +9,7 @@ tags:
 - xlsr-fine-tuning-week
 license: apache-2.0
 model-index:
-- name: wav2vec2-large-xlsr-53-polish by Alex Leu
+- name: XLSR Wav2vec2 Large 53 Polish by Alex Leu
   results:
   - task: 
       name: Speech Recognition
@@ -21,8 +21,9 @@ model-index:
     metrics:
        - name: Test WER
          type: wer
-         value:xxx
+         value: xxx
 ---
+
 # wav2vec2-large-xlsr-polish
 Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) in Polish using the [Common Voice](https://huggingface.co/datasets/common_voice)
 When using this model, make sure that your speech input is sampled at 16kHz.
@@ -64,7 +65,7 @@ wer = load_metric("wer")
 processor = Wav2Vec2Processor.from_pretrained("alexcleu/wav2vec2-large-xlsr-polish")
 model = Wav2Vec2ForCTC.from_pretrained("alexcleu/wav2vec2-large-xlsr-polish")
 model.to("cuda")
-chars_to_ignore_regex = '[\\,\\?\\.\\!\\-\\;\\:\\"\\“]'
+chars_to_ignore_regex = '[\\\\\\\\\\\\\\\\,\\\\\\\\\\\\\\\\?\\\\\\\\\\\\\\\\.\\\\\\\\\\\\\\\\!\\\\\\\\\\\\\\\\-\\\\\\\\\\\\\\\\;\\\\\\\\\\\\\\\\:\\\\\\\\\\\\\\\\"\\\\\\\\\\\\\\\\“]'
 resampler = torchaudio.transforms.Resample(48_000, 16_000)
 # Preprocessing the datasets.
 # We need to read the aduio files as arrays