Update handler.py

handler.py

@@ -11,35 +11,35 @@ WhisperProcessor
 from peft import LoraConfig, PeftModel, LoraModel, LoraConfig, get_peft_model, PeftConfig
 
 class EndpointHandler():
-def __init__(self, path=""):
-    # self.pipeline = pipeline(task= "automatic-speech-recognition", model=path)
-    # self.pipeline.model.config.forced_decoder_ids = self.pipeline.tokenizer.get_decoder_prompt_ids(language="Chinese", task="transcribe")
-    # self.pipeline.model.generation_config.forced_decoder_ids = self.pipeline.model.config.forced_decoder_ids  # just to be sure!
-    peft_model_id = "cathyi/openai-whisper-large-v2-Lora"
-    language = "Chinese"
-    task = "transcribe"
-    peft_config = PeftConfig.from_pretrained(peft_model_id)
-    model = WhisperForConditionalGeneration.from_pretrained(
-        peft_config.base_model_name_or_path
-    )
-    model = PeftModel.from_pretrained(model, peft_model_id)
-    tokenizer = WhisperTokenizer.from_pretrained(peft_config.base_model_name_or_path, language=language, task=task)
-    processor = WhisperProcessor.from_pretrained(peft_config.base_model_name_or_path, language=language, task=task)
-    feature_extractor = processor.feature_extractor
-    self.forced_decoder_ids = processor.get_decoder_prompt_ids(language=language, task=task)
-    pipeline = AutomaticSpeechRecognitionPipeline(model=model, tokenizer=tokenizer, feature_extractor=feature_extractor)
-
-def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]:
-    """
-    data args:
-        inputs (:obj: `str`)
-        date (:obj: `str`)
-    Return:
-        A :obj:`list` | `dict`: will be serialized and returned
-    """
-    # get inputs
+    def __init__(self, path=""):
+        # self.pipeline = pipeline(task= "automatic-speech-recognition", model=path)
+        # self.pipeline.model.config.forced_decoder_ids = self.pipeline.tokenizer.get_decoder_prompt_ids(language="Chinese", task="transcribe")
+        # self.pipeline.model.generation_config.forced_decoder_ids = self.pipeline.model.config.forced_decoder_ids  # just to be sure!
+        peft_model_id = "cathyi/openai-whisper-large-v2-Lora"
+        language = "Chinese"
+        task = "transcribe"
+        peft_config = PeftConfig.from_pretrained(path)
+        model = WhisperForConditionalGeneration.from_pretrained(
+            peft_config.base_model_name_or_path
+        )
+        model = PeftModel.from_pretrained(model, peft_model_id)
+        tokenizer = WhisperTokenizer.from_pretrained(peft_config.base_model_name_or_path, language=language, task=task)
+        processor = WhisperProcessor.from_pretrained(peft_config.base_model_name_or_path, language=language, task=task)
+        feature_extractor = processor.feature_extractor
+        self.forced_decoder_ids = processor.get_decoder_prompt_ids(language=language, task=task)
+        pipeline = AutomaticSpeechRecognitionPipeline(model=model, tokenizer=tokenizer, feature_extractor=feature_extractor)
     
-    # run normal prediction
-    inputs = data.pop("inputs", data)
-    prediction = self.pipeline(inputs, generate_kwargs={"forced_decoder_ids": self.forced_decoder_ids}, max_new_tokens=255)["text"]
-    return prediction
+    def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]:
+        """
+        data args:
+            inputs (:obj: `str`)
+            date (:obj: `str`)
+        Return:
+            A :obj:`list` | `dict`: will be serialized and returned
+        """
+        # get inputs
+        
+        # run normal prediction
+        inputs = data.pop("inputs", data)
+        prediction = self.pipeline(inputs, generate_kwargs={"forced_decoder_ids": self.forced_decoder_ids}, max_new_tokens=255)["text"]
+        return prediction