---
language: bem #TODO: replace {lang_id} in your language code here. Make sure the code is one of the *ISO codes* of [this](https://huggingface.co/languages) site.
datasets:
- BembaSpeech #TODO: remove if you did not use the common voice dataset
#- TODO: add more datasets if you have used additional datasets. Make sure to use the exact same dataset name as the one found [here](https://huggingface.co/datasets). If the dataset can not be found in the official datasets, just give it a new name
metrics:
- wer
tags:
- audio
- automatic-speech-recognition
- speech
- xlsr-fine-tuning-week
license: apache-2.0
model-index:
- name: XLSR Wav2Vec2 Bemba by Claytone Sikasote     #TODO: replace {human_readable_name} with a name of your model as it should appear on the leaderboard. It could be something like `Elgeish XLSR Wav2Vec2 Large 53`
  results:
  - task: 
      name: Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: BembaSpeech bem #TODO: replace {lang_id} in your language code here. Make sure the code is one of the *ISO codes* of [this](https://huggingface.co/languages) site.
      type: bembaspeech
      args: bem #TODO: replace {lang_id} in your language code here. Make sure the code is one of the *ISO codes* of [this](https://huggingface.co/languages) site.
    metrics:
       - name: Test WER
         type: wer
         value: 42.14 #TODO (IMPORTANT): replace {wer_result_on_test} with the WER error rate you achieved on the common_voice test set. It should be in the format XX.XX (don't add the % sign here). **Please** remember to fill out this value after you evaluated your model, so that your model appears on the leaderboard. If you fill out this model card before evaluating your model, please remember to edit the model card afterward to fill in your value
---

# Wav2Vec2-Large-XLSR-53-Bemba #TODO: replace language with your {language}, *e.g.* French

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). #TODO: replace {language} with your language, *e.g.* French and eventually add more datasets that were used and eventually remove common voice if model was not trained on common voice
When using this model, make sure that your speech input is sampled at 16kHz.

## Usage

The model can be used directly (without a language model) as follows:

```python
import torch
import torchaudio
from datasets import load_dataset
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor

test_dataset = load_dataset("common_voice", "{lang_id}", split="test[:2%]") #TODO: replace {lang_id} in your language code here. Make sure the code is one of the *ISO codes* of [this](https://huggingface.co/languages) site.

processor = Wav2Vec2Processor.from_pretrained("{model_id}") #TODO: replace {model_id} with your model id. The model id consists of {your_username}/{your_modelname}, *e.g.* `elgeish/wav2vec2-large-xlsr-53-arabic`
model = Wav2Vec2ForCTC.from_pretrained("{model_id}") #TODO: replace {model_id} with your model id. The model id consists of {your_username}/{your_modelname}, *e.g.* `elgeish/wav2vec2-large-xlsr-53-arabic`

resampler = torchaudio.transforms.Resample(48_000, 16_000)

# Preprocessing the datasets.
# We need to read the aduio files as arrays
def speech_file_to_array_fn(batch):
	speech_array, sampling_rate = torchaudio.load(batch["path"])
	batch["speech"] = resampler(speech_array).squeeze().numpy()
	return batch

test_dataset = test_dataset.map(speech_file_to_array_fn)
inputs = processor(test_dataset["speech"][:2], sampling_rate=16_000, return_tensors="pt", padding=True)

with torch.no_grad():
	logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits

predicted_ids = torch.argmax(logits, dim=-1)

print("Prediction:", processor.batch_decode(predicted_ids))
print("Reference:", test_dataset["sentence"][:2])
```


## Evaluation

The model can be evaluated as follows on the Bemba test data of BembaSpeech.  # TODO: replace #TODO: replace language with your {language}, *e.g.* French


```python
import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import re

test_dataset = load_dataset("common_voice", "{lang_id}", split="test") #TODO: replace {lang_id} in your language code here. Make sure the code is one of the *ISO codes* of [this](https://huggingface.co/languages) site.
wer = load_metric("wer")

processor = Wav2Vec2Processor.from_pretrained("{model_id}") #TODO: replace {model_id} with your model id. The model id consists of {your_username}/{your_modelname}, *e.g.* `elgeish/wav2vec2-large-xlsr-53-arabic`
model = Wav2Vec2ForCTC.from_pretrained("{model_id}") #TODO: replace {model_id} with your model id. The model id consists of {your_username}/{your_modelname}, *e.g.* `elgeish/wav2vec2-large-xlsr-53-arabic`
model.to("cuda")

chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“]'  # TODO: adapt this list to include all special characters you removed from the data
#resampler = torchaudio.transforms.Resample(48_000, 16_000)

# Preprocessing the datasets.
# We need to read the aduio files as arrays
def speech_file_to_array_fn(batch):
	batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
	speech_array, sampling_rate = torchaudio.load(batch["path"])
	batch["speech"] = resampler(speech_array).squeeze().numpy()
	return batch

test_dataset = test_dataset.map(speech_file_to_array_fn)

# Preprocessing the datasets.
# We need to read the aduio files as arrays
def evaluate(batch):
	inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)

	with torch.no_grad():
		logits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits

	pred_ids = torch.argmax(logits, dim=-1)
	batch["pred_strings"] = processor.batch_decode(pred_ids)
	return batch

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 %  # TODO: write output of print here. IMPORTANT: Please remember to also replace {wer_result_on_test} at the top of with this value here. tags.


## Training

The BembaSpeech `train`, `dev` and `test` datasets were used for training, development and evaluation respectively  # TODO: adapt to state all the datasets that were used for training.

The script used for training can be found [here](https://colab.research.google.com/drive/1IgdR-EQq5rgmBqw5O6tcfJpmXM8rDX55?usp=sharing) # TODO: fill in a link to your training script here. If you trained your model in a colab, simply fill in the link here. If you trained the model locally, it would be great if you could upload the training script on github and paste the link here.