Update app.py

app.py

@@ -13,22 +13,9 @@ title1 = "RWKV-4-Raven-7B-v8-Eng-20230408-ctx4096"
 os.environ["RWKV_JIT_ON"] = '1'
 os.environ["RWKV_CUDA_ON"] = '1' # if '1' then use CUDA kernel for seq mode (much faster)
 
-os.system('pip install voicefixer --upgrade')
-from voicefixer import VoiceFixer
-voicefixer = VoiceFixer()
-
 from TTS.api import TTS
 tts = TTS(model_name="tts_models/multilingual/multi-dataset/your_tts", progress_bar=False, gpu=True)
 
-import torchaudio
-from speechbrain.pretrained import SpectralMaskEnhancement
-
-enhance_model = SpectralMaskEnhancement.from_hparams(
-source="speechbrain/metricgan-plus-voicebank",
-savedir="pretrained_models/metricgan-plus-voicebank",
-run_opts={"device":"cuda"},
-)
-
 from rwkv.model import RWKV
 model_path = hf_hub_download(repo_id="BlinkDL/rwkv-4-raven", filename=f"{title1}.pth")
 model = RWKV(model=model_path, strategy='cuda fp16i8 *8 -> cuda fp16')
@@ -122,10 +109,12 @@ def evaluate(
 
     res1 = ' '.join(str(x) for x in res)
     
-
-    return [result.text, res]
+    tts.tts_to_file(res1, speaker_wav = upload, language="en", file_path="output.wav")
+    
+#    return [result.text, res]
+    
+    return [result.text, res, "output.wav"]    
     
-#    return [result.text, res, "enhanced.wav"]    
 #    yield out_str.strip()
 
 g = gr.Interface(
@@ -150,7 +139,7 @@ g = gr.Interface(
             lines=5,
             label="Raven Output",
         ),
-#        gr.Audio(label="Audio with Custom Voice"),
+        gr.Audio(label="Audio with Custom Voice"),
     ],
     title="🥳💬💕 - TalktoAI，随时随地，谈天说地！",
     description="🤖 - 让有人文关怀的AI造福每一个人！AI向善，文明璀璨！TalktoAI - Enable the future！",