(wip)debug

tts.py

@@ -7,6 +7,8 @@ import fal_client
 import requests
 import time
 import io
+
+from gradio_client import handle_file
 from pyht import Client as PyhtClient
 from pyht.client import TTSOptions
 import base64
@@ -204,8 +206,10 @@ def predict_index_tts(text, reference_audio_path=None):
     result = client.predict(
         prompt=prompt,
         text=text,
+        stream="false",
         api_name="/gen_single"
     )
+    print("index-tts result:", result)
     return result
 
 
@@ -220,8 +224,10 @@ def predict_spark_tts(text, reference_audio_path=None):
         prompt_text=text,
         prompt_wav_upload=prompt_wav,
         prompt_wav_record=prompt_wav,
+        stream="false",
         api_name="/voice_clone"
     )
+    print("spark-tts result:", result)
     return result
 
 
@@ -230,7 +236,7 @@ def predict_cosyvoice_tts(text, reference_audio_path=None):
     client = Client("https://iic-cosyvoice2-0-5b.ms.show/")
     if not reference_audio_path:
         raise ValueError("cosyvoice-2.0 需要 reference_audio_path")
-    prompt_wav = file(reference_audio_path)
+    prompt_wav = handle_file(reference_audio_path)
     result = client.predict(
         tts_text=text,
         mode_checkbox_group="3s极速复刻",
@@ -242,6 +248,8 @@ def predict_cosyvoice_tts(text, reference_audio_path=None):
         stream="false",
         api_name="/generate_audio"
     )
+    print("cosyvoice-2.0 result:", result)
+
     return result