Update app.py

app.py

@@ -1,33 +1,11 @@
+import torch
+from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
+from datasets import load_dataset
 from fastapi import FastAPI, Request, HTTPException
-from transformers import pipeline
-import io
 import librosa
-from transformers import WhisperForConditionalGeneration, WhisperProcessor
-
+import io
+whisper = pipeline("automatic-speech-recognition", "whitefox123/whisper-small-ar2", torch_dtype=torch.float16, device="cpu")
 app = FastAPI()
-# Device configuration
-# Load the model and processor
-model_id = "whitefox123/whisper-small-ar2"
-model = WhisperForConditionalGeneration.from_pretrained(
-    model_id
-)
-import torch
-
-processor = WhisperProcessor.from_pretrained('whitefox123/whisper-small-ar2')
-model.config.forced_decoder_ids = None
-forced_decoder_ids = processor.get_decoder_prompt_ids(language="Arabic", task="transcribe")
-model.generation_config.cache_implementation = "static"
-from transformers import GenerationConfig, WhisperForConditionalGeneration
-generation_config = GenerationConfig.from_pretrained("openai/whisper-small") # if you are using a multilingual model
-model.generation_config = generation_config
-
-pipe = pipeline(
-    "automatic-speech-recognition",
-    model=model,
-    tokenizer=processor.tokenizer,
-    feature_extractor=processor.feature_extractor,
-    
-)
 
 @app.post("/transcribe/")
 async def transcribe_audio(request: Request):
@@ -41,18 +19,13 @@ async def transcribe_audio(request: Request):
         # Load the audio file using pydub
         audio_array, sampling_rate = librosa.load(audio_file, sr=16000)
 
-        # Process the audio array
-        input_features = processor(audio_array, sampling_rate=sampling_rate, return_tensors="pt").input_features 
-
-        # Generate token ids
-        predicted_ids = model.generate(input_features, forced_decoder_ids=forced_decoder_ids, return_timestamps=True)
-
         # Decode token ids to text
-        transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
+        transcription = whisper(audio_array)
+
 
         # Print the transcription
-        print(transcription[0])  # Display the transcriptiontry:
+        print(transcription['text'])  # Display the transcriptiontry:
        
-        return {"transcription": transcription[0]}
+        return {"transcription": transcription['text']}
     except Exception as e:
         raise HTTPException(status_code=500, detail=str(e))