muzamil47
/

wav2vec2-large-xlsr-53-arabic-demo

Automatic Speech Recognition

xlsr-fine-tuning-week

Inference Endpoints

Model card Files Files and versions Community

muzamil47 commited on Dec 21, 2022

Commit

38e2413

•

1 Parent(s): 0ca57f5

Update README.md

Files changed (1) hide show

README.md +0 -2

README.md CHANGED Viewed

@@ -70,7 +70,6 @@ def predict(data):
         predicted = torch.argmax(model(input_values, attention_mask=attention_mask).logits, dim=-1)
     data["predicted"] = processor.tokenizer.decode(predicted[0])
-    print(data["predicted"])
     print("predicted:", buckwalter.untrans(data["predicted"]))
     return data
@@ -78,7 +77,6 @@ predict(load_file_to_data("common_voice_ar_19058307.mp3"))
 ```
 **Output Result**:
 ```shell
-reference: هل يمكنني التحدث مع المسؤول هنا
 predicted: هل يمكنني التحدث مع المسؤول هنا
 ```

         predicted = torch.argmax(model(input_values, attention_mask=attention_mask).logits, dim=-1)
     data["predicted"] = processor.tokenizer.decode(predicted[0])
     print("predicted:", buckwalter.untrans(data["predicted"]))
     return data
 ```
 **Output Result**:
 ```shell
 predicted: هل يمكنني التحدث مع المسؤول هنا
 ```