fixed typo in speech encoder decoder doc (#25745)

fixed typo in speech encoder decoder blog

docs/source/en/model_doc/speech-encoder-decoder.md

@@ -111,7 +111,7 @@ speech inputs) and `labels` (which are the `input_ids` of the encoded target seq
 >>> labels = tokenizer(ds[0]["text"], return_tensors="pt").input_ids
 
 >>> # the forward function automatically creates the correct decoder_input_ids
->>> loss = model(**input_features).loss
+>>> loss = model(input_values=input_values, labels=labels).loss
 >>> loss.backward()
 ```
 
@@ -129,4 +129,4 @@ speech inputs) and `labels` (which are the `input_ids` of the encoded target seq
 
 [[autodoc]] FlaxSpeechEncoderDecoderModel
     - __call__
-    - from_encoder_decoder_pretrained
+    - from_encoder_decoder_pretrained
\ No newline at end of file