Update custom model files, README, and requirements

asr_pipeline.py

@@ -42,7 +42,9 @@ class ASRPipeline(transformers.AutomaticSpeechRecognitionPipeline):
             yield item
 
     def _forward(self, model_inputs, **generate_kwargs) -> dict[str, Any]:
-        # Extract audio features
+        # Extract audio features and is_last flag
+        is_last = model_inputs.pop("is_last", True) if isinstance(model_inputs, dict) else True
+
         if isinstance(model_inputs, dict):
             input_features = model_inputs.get("input_features")
             if input_features is not None:
@@ -55,7 +57,7 @@ class ASRPipeline(transformers.AutomaticSpeechRecognitionPipeline):
             **generate_kwargs,
         )
 
-        return {"tokens": generated_ids}
+        return {"tokens": generated_ids, "is_last": is_last}
 
     def postprocess(self, model_outputs, **kwargs) -> dict[str, str]:
         tokens = model_outputs.get("tokens")