Update README.md

README.md

@@ -17,11 +17,11 @@ model-index:
     metrics:
        - name: Test WER
          type: wer
-         value: 49.68
+         value: 44.30
 ---
 # Wav2Vec2-Large-XLSR-53-Moroccan-Darija
 
-**wav2vec2-large-xlsr-53** fine-tuned on 6 hours of labeled Darija Audios
+**wav2vec2-large-xlsr-53** fine-tuned on 8.5 hours of labeled Darija Audios
 
 I have also added 3 phonetic units to this model ڭ, ڤ and پ. For example: ڭال , ڤيديو , پودكاست
 
@@ -59,7 +59,17 @@ print(transcription)
 
 Here's the output: ڭالت ليا هاد السيد هادا ما كاينش بحالو 
 
-## Evaluation
+## Evaluation & Previous works
+
+-v2 (fine-tuned on 8.5 hours of audio + replacing أ and ى and إ with ا + tried to standardize the Moroccan Darija)
+
+**Wer**: 44.30
+
+**Training Loss**: 12.99
+
+**Validation Loss**: 36.93
+
+-v1 (fine-tuned on 6 hours of audio)
 
 **Wer**: 49.68
 
@@ -67,10 +77,10 @@ Here's the output: ڭالت ليا هاد السيد هادا ما كاينش ب
 
 **Validation Loss**: 45.24
 
-This high validation loss value is mainly due to the fact that Darija can be written in many ways.
-
 ## Future Work
 
-Currently working on improving this model. The new model will be available soon.
+I am currently working on improving this model. The new model will be available soon.
+
+email: souregh@gmail.com