upd README.md

README.md

@@ -1,9 +1,7 @@
 ---
 language: cs
-datasets:
-- common_voice
-metrics:
-- wer
+dataset: common_voice
+metrics: wer
 tags:
 - audio
 - automatic-speech-recognition
@@ -11,7 +9,7 @@ tags:
 - xlsr-fine-tuning-week
 license: apache-2.0
 model-index:
-- name: `Czech XLSR Wav2Vec2 Large 53`
+- name: Czech XLSR Wav2Vec2 Large 53
   results:
   - task: 
       name: Speech Recognition
@@ -26,7 +24,7 @@ model-index:
          value: 24.93
 ---
 
-# Wav2Vec2-Large-XLSR-53-{language} #TODO: replace language with your {language}, *e.g.* French
+# Wav2Vec2-Large-XLSR-53-Chech
 
 Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on Czech using the [Common Voice](https://huggingface.co/datasets/common_voice) dataset.
 When using this model, make sure that your speech input is sampled at 16kHz.
@@ -50,15 +48,15 @@ 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
+\tspeech_array, sampling_rate = torchaudio.load(batch["path"])
+\tbatch["speech"] = resampler(speech_array).squeeze().numpy()
+\treturn 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
+\tlogits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
 
 predicted_ids = torch.argmax(logits, dim=-1)
 
@@ -94,30 +92,30 @@ resampler = torchaudio.transforms.Resample(48_000, 16_000)
 # We need to read the aduio files as arrays
 # Note: this models is trained ignoring accents on letters as below
 def speech_file_to_array_fn(batch):
-	batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().strip()
-	batch["sentence"] = re.sub(re.compile('[äá]'), 'a', batch['sentence'])
-	batch["sentence"] = re.sub(re.compile('[öó]'), 'o', batch['sentence'])
-	batch["sentence"] = re.sub(re.compile('[èé]'), 'e', batch['sentence'])
-	batch["sentence"] = re.sub(re.compile("[ïí]"), 'i', batch['sentence'])
-	batch["sentence"] = re.sub(re.compile("[üů]"), 'u', batch['sentence'])
-	batch['sentence'] = re.sub('  ', ' ', batch['sentence'])
-	speech_array, sampling_rate = torchaudio.load(batch["path"])
-	batch["speech"] = resampler(speech_array).squeeze().numpy()
-	return batch
+\tbatch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().strip()
+\tbatch["sentence"] = re.sub(re.compile('[äá]'), 'a', batch['sentence'])
+\tbatch["sentence"] = re.sub(re.compile('[öó]'), 'o', batch['sentence'])
+\tbatch["sentence"] = re.sub(re.compile('[èé]'), 'e', batch['sentence'])
+\tbatch["sentence"] = re.sub(re.compile("[ïí]"), 'i', batch['sentence'])
+\tbatch["sentence"] = re.sub(re.compile("[üů]"), 'u', batch['sentence'])
+\tbatch['sentence'] = re.sub('  ', ' ', batch['sentence'])
+\tspeech_array, sampling_rate = torchaudio.load(batch["path"])
+\tbatch["speech"] = resampler(speech_array).squeeze().numpy()
+\treturn 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)
+\tinputs = 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
+\twith torch.no_grad():
+\t\tlogits = 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
+\tpred_ids = torch.argmax(logits, dim=-1)
+\tbatch["pred_strings"] = processor.batch_decode(pred_ids)
+\treturn batch
 
 result = test_dataset.map(evaluate, batched=True, batch_size=8)