Update app.py

app.py

@@ -4,7 +4,7 @@ os.system("pip install transformers==4.27.0")
 os.system("pip install torch")
 os.system("pip install openai")
 os.system("pip install accelerate")
-from transformers import pipeline, WhisperModel, WhisperTokenizer, WhisperFeatureExtractor, AutoFeatureExtractor, AutoProcessor
+from transformers import pipeline, WhisperModel, WhisperTokenizer, WhisperFeatureExtractor, AutoFeatureExtractor, AutoProcessor, WhisperConfig
 os.system("pip install evaluate")
 #import evaluate
 #os.system("pip install evaluate[evaluator]")
@@ -27,17 +27,10 @@ huggingface_token = os.environ["huggingface_token"]
 model = WhisperModel.from_pretrained("mskov/whisper_miso", use_auth_token=huggingface_token)
 feature_extractor = AutoFeatureExtractor.from_pretrained("mskov/whisper_miso", use_auth_token=huggingface_token)
 
+model_config = WhisperConfig.from_pretrained("mskov/whisper_miso", use_auth_token=huggingface_token)
+model = WhisperModel(config=model_config)
+model.load_state_dict(torch.load("mskov/whisper_miso/checkpoint-4000/pytorch_model.bin"))
+model.eval()
 
-ds = load_dataset("mskov/miso_test", split="test")
-print("testing testing")
-ds = ds.cast_column("audio", Audio(sampling_rate=16000))
-
-
-inputs = feature_extractor(ds[0]["audio"]["array"], return_tensors="pt")
-print("check check")
-print(inputs)
-input_features = inputs.input_features
-decoder_input_ids = torch.tensor([[1, 1]]) * model.config.decoder_start_token_id
-last_hidden_state = model(input_features, decoder_input_ids=decoder_input_ids).last_hidden_state
-list(last_hidden_state.shape)
-print(list(last_hidden_state.shape))
+dataset = load_dataset("mskov/miso_test", split="test").cast_column("audio", Audio())
+print(dataset)