Update README.md

README.md

@@ -23,12 +23,12 @@ model-index:
     metrics:
        - name: Test WER
          type: wer
-         value: 42.14 
+         value: 42.17 
 ---
 
 # Wav2Vec2-Large-XLSR-53-Bemba
 
-Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on Bemba using the [BembaSpeech](https://csikasote.github.io/BembaSpeech). When using this model, make sure that your speech input is sampled at 16kHz.
+Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on Bemba language of Zambia using the [BembaSpeech](https://csikasote.github.io/BembaSpeech). When using this model, make sure that your speech input is sampled at 16kHz.
 
 ## Usage
 
@@ -79,14 +79,14 @@ from datasets import load_dataset, load_metric
 from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
 import re
 
-test_dataset = load_dataset("csv", data_files={"test": "/content/test.csv"}, delimiter="\t")["test"] 
+test_dataset = load_dataset("csv", data_files={"test": "/content/test.csv"}, delimiter="\\t")["test"] 
 wer = load_metric("wer")
 
 processor = Wav2Vec2Processor.from_pretrained("csikasote/wav2vec2-large-xlsr-bemba") 
 model = Wav2Vec2ForCTC.from_pretrained("csikasote/wav2vec2-large-xlsr-bemba") 
 model.to("cuda")
 
-chars_to_ignore_regex = '[\,\?\.\!\;\:\"\“]'  
+chars_to_ignore_regex = '[\,\_\?\.\!\;\:\"\“]'  
 #resampler = torchaudio.transforms.Resample(48_000, 16_000)
 
 # Preprocessing the datasets.
@@ -116,8 +116,8 @@ result = test_dataset.map(evaluate, batched=True, batch_size=8)
 print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
 ```
 
-**Test Result**: 42.14 %  
+**Test Result**: 42.17 %  
 
 ## Training
 
-The BembaSpeech `train`, `dev` and `test` datasets were used for training, development and evaluation respectively. The script used for training can be found [here](https://colab.research.google.com/drive/1IgdR-EQq5rgmBqw5O6tcfJpmXM8rDX55?usp=sharing).
+The BembaSpeech `train`, `dev` and `test` datasets were used for training, development and evaluation respectively. The script used for evaluating the model on the test dataset can be found [here](https://colab.research.google.com/drive/1aplFHfaXE68HGDwBYV2KqUWPasrk7bXv?usp=sharing).