feat:clean up playground explore audio preprocess, todo:build free run from pure input data

.gitignore

@@ -193,6 +193,8 @@ cython_debug/
 
 # PyPI configuration file
 .pypirc
+
 onnx/
 *.onnx
 tokenizer.json
+output.wav

playground/export_hubert.py

@@ -67,7 +67,7 @@ class HubertONNXExporter:
             print(f"[Error] ONNX model not found at {self.onnx_path}")
             return False
             
-    def _load_and_preprocess_audio(self, audio_path, max_length=32000):
+    def _load_and_preprocess_audio(self, audio_path, max_length=160000):
         """Load and preprocess audio file"""
         waveform, sample_rate = torchaudio.load(audio_path)
         
@@ -80,10 +80,18 @@ class HubertONNXExporter:
         if waveform.shape[0] > 1:
             waveform = waveform[0:1]
         
-        # Limit length for testing (2 seconds at 16kHz)
+        # Limit length for testing (10 seconds at 16kHz)
         if waveform.shape[1] > max_length:
             waveform = waveform[:, :max_length]
 
+        # make a zero tensor that has length 3200*0.3
+        zero_tensor = torch.zeros((1, 9600), dtype=torch.float32)
+
+        print("waveform shape and zero wave shape", waveform.shape, zero_tensor.shape)
+
+        # concate zero_tensor with waveform
+        waveform = torch.cat([waveform, zero_tensor], dim=1)
+
         return waveform
         
     def test_torch_vs_onnx(self, audio_path="playground/ref/audio.wav"):

playground/freerun.py

@@ -29,17 +29,54 @@ def audio_postprocess(
 
     return audio
 
+def load_and_preprocess_audio(audio_path, max_length=160000):
+    """Load and preprocess audio file"""
+    waveform, sample_rate = torchaudio.load(audio_path)
     
-input_phones = np.load("playground/input_phones.npy")
+    # Resample to 16kHz if needed
-input_bert = np.load("playground/input_bert.npy").T.astype(np.float32)
+    if sample_rate != 16000:
-ref_phones = np.load("playground/ref_phones.npy")
+        resampler = torchaudio.transforms.Resample(sample_rate, 16000)
-ref_bert = np.load("playground/ref_bert.npy").T.astype(np.float32)
+        waveform = resampler(waveform)
-audio_prompt_hubert = np.load("playground/audio_prompt_hubert.npy").astype(np.float32)
+    
+    # Take first channel
+    if waveform.shape[0] > 1:
+        waveform = waveform[0:1]
+    
+    # Limit length for testing (10 seconds at 16kHz)
+    if waveform.shape[1] > max_length:
+        waveform = waveform[:, :max_length]
+
+    # make a zero tensor that has length 3200*0.3
+    zero_tensor = torch.zeros((1, 9600), dtype=torch.float32)
+
+    print("waveform shape and zero wave shape", waveform.shape, zero_tensor.shape)
+
+    # concate zero_tensor with waveform
+    waveform = torch.cat([waveform, zero_tensor], dim=1)
+
+    return waveform
+
+def get_audio_hubert(audio_path):
+    """Get HuBERT features for the audio file"""
+    waveform = load_and_preprocess_audio(audio_path)
+    ort_session = ort.InferenceSession("playground/hubert/chinese-hubert-base.onnx")
+    ort_inputs = {ort_session.get_inputs()[0].name: waveform.numpy()}
+    hubert_feature = ort_session.run(None, ort_inputs)[0].astype(np.float32)
+    # transpose axis 1 and 2 with numpy
+    hubert_feature = hubert_feature.transpose(0, 2, 1)
+    print("Hubert feature shape:", hubert_feature.shape)
+    return hubert_feature
+
+input_phones = np.load("playground/ref/input_phones.npy")
+input_bert = np.load("playground/ref/input_bert.npy").T.astype(np.float32)
+ref_phones = np.load("playground/ref/ref_phones.npy")
+ref_bert = np.load("playground/ref/ref_bert.npy").T.astype(np.float32)
+audio_prompt_hubert = get_audio_hubert("playground/ref/audio.wav")
 
 
 encoder = ort.InferenceSession(MODEL_PATH+"_export_t2s_encoder.onnx")
 
-outputs = encoder.run(None, {
+[x, prompts] = encoder.run(None, {
     "text_seq": input_phones,
     "text_bert": input_bert,
     "ref_seq": ref_phones,
@@ -47,10 +84,7 @@ outputs = encoder.run(None, {
     "ssl_content": audio_prompt_hubert
 })
 
-print(outputs[0].shape, outputs[1].shape)
+print(x.shape, prompts.shape)
-
-x = outputs[0]
-prompts = outputs[1]
 
 fsdec = ort.InferenceSession(MODEL_PATH+"_export_t2s_fsdec.onnx")
 sdec = ort.InferenceSession(MODEL_PATH+"_export_t2s_sdec.onnx")