first update

README.md

@@ -1,13 +1,11 @@
 ---
-title: Tts Ru Free Hf Vits Low Multispeaker
-emoji: 😻
-colorFrom: gray
-colorTo: gray
+title: Text to Speech Russian free multispeaker model
+emoji: 🐠
+colorFrom: purple
+colorTo: yellow
 sdk: gradio
-sdk_version: 4.28.3
+sdk_version: 4.19.2
 app_file: app.py
 pinned: false
-license: mit
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,40 @@
+import gradio as gr
+from transformers import VitsModel, AutoTokenizer, set_seed
+import torch
+import scipy.io.wavfile
+from ruaccent import RUAccent
+
+speakers={"man": 0,
+        "woman": 1}
+
+model = VitsModel.from_pretrained("utrobinmv/tts_ru_free_hf_vits_low_multispeaker")
+tokenizer = AutoTokenizer.from_pretrained("utrobinmv/tts_ru_free_hf_vits_low_multispeaker")
+model.eval()
+set_seed(555)
+
+accentizer = RUAccent()
+accentizer.load(omograph_model_size='turbo', use_dictionary=True)
+
+
+def generate_audio(speaker_name, text):
+    text = accentizer.process_all(text)
+    inputs = tokenizer(text, return_tensors="pt")
+    inputs['speaker_id'] = speakers[speaker_name]
+    
+    with torch.no_grad():
+        output = model(**inputs).waveform
+        
+    scipy.io.wavfile.write("output.wav", rate=model.config.sampling_rate, data=output[0].cpu().numpy())
+    
+    return "output.wav"
+
+speaker_dropdown = gr.Dropdown(
+            choices=speakers,
+            label="Speaker id",
+            value='woman',
+            info=f"Models are trained on 2 speakers",  
+            interactive=True   
+        )
+
+iface = gr.Interface(fn=generate_audio, inputs=[speaker_dropdown,"text"], outputs="audio", title="Text to Speech Russian free multispeaker model", description="Введите текст на русском языке, чтобы преобразовать его в русскую звуковую речь. Пример текста: Привет, как дела? А у тебя как?")
+iface.launch(share=True)

requirements.txt

@@ -0,0 +1,5 @@
+torch
+scipy
+sentencepiece==0.1.99
+ruaccent==1.5.6.3
+transformers==4.38.1