Update README.md

README.md

@@ -77,6 +77,14 @@ print(prediction[0])
 
 ## Evaluation
 For the evaluation use the code below:
+```
+*Input csv files format:*
+
+| path| reference|
+|---|---|
+| path to audio files | corresponding transcription|
+
+```
 ```python
 import torch
 import torchaudio
@@ -113,14 +121,6 @@ def predict(batch):
     return batch
     
 dataset = load_dataset("csv", data_files={"test":"path/to/your.csv"}, delimiter=",")["test"] 
-```
-*Input csv files format:*
-
-| path| reference|
-|---|---|
-| path to audio files | corresponding transcription|
-
-```
 dataset = dataset.map(speech_file_to_array_fn)
 
 result = dataset.map(predict, batched=True, batch_size=4)