mod chunk 10->15

app.py

@@ -14,7 +14,7 @@ from parler_tts import ParlerTTSForConditionalGeneration
 from pydub import AudioSegment
 from transformers import AutoTokenizer, AutoFeatureExtractor, set_seed
 
-device = "cuda:0" if torch.cuda.is_available() else "mps" if torch.backends.mps.is_available() else "cpu"
+device = "cuda" if torch.cuda.is_available() else "mps" if torch.backends.mps.is_available() else "cpu"
 torch_dtype = torch.bfloat16 if device != "cpu" else torch.float32
 
 repo_id = "ai4bharat/indic-parler-tts-pretrained"
@@ -200,7 +200,7 @@ frame_rate = model.audio_encoder.config.frame_rate
 def generate_base(text, description, play_steps_in_s=2.0):
     # Initialize variables
     play_steps = int(frame_rate * play_steps_in_s)
-    chunk_size = 10  # Process 10 words at a time
+    chunk_size = 15  # Process 10 words at a time
     
     # Tokenize the full text and description
     inputs = description_tokenizer(description, return_tensors="pt").to(device)
@@ -272,7 +272,7 @@ def generate_base(text, description, play_steps_in_s=2.0):
 def generate_jenny(text, description, play_steps_in_s=2.0):
     # Initialize variables
     play_steps = int(frame_rate * play_steps_in_s)
-    chunk_size = 10  # Process 10 words at a time
+    chunk_size = 15  # Process 10 words at a time
     
     # Tokenize the full text and description
     inputs = description_tokenizer(description, return_tensors="pt").to(device)