edwko commited on
Commit
e6297e8
·
verified ·
1 Parent(s): 997b839

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +9 -7
README.md CHANGED
@@ -117,12 +117,7 @@ model_config = outetts.HFModelConfig_v2(
117
  interface = outetts.InterfaceHF(model_version="0.3", cfg=model_config)
118
 
119
  # You can create a speaker profile for voice cloning, which is compatible across all backends.
120
- # speaker = interface.create_speaker(
121
- # audio_path="path/to/audio/file.wav",
122
- # transcript=None, # Set to None to use Whisper for transcription
123
- # whisper_model="turbo", # Optional: specify Whisper model (default: "turbo")
124
- # whisper_device=None, # Optional: specify device for Whisper (default: None)
125
- # )
126
  # interface.save_speaker(speaker, "speaker.json")
127
  # speaker = interface.load_speaker("speaker.json")
128
 
@@ -144,8 +139,15 @@ output = interface.generate(config=gen_cfg)
144
  # Save the generated speech to a file
145
  output.save("output.wav")
146
  ```
 
 
 
 
 
 
 
147
  > [!IMPORTANT]
148
- > ## For additional usage examples and recommendations, visit the [GitHub repository](https://github.com/edwko/OuteTTS?tab=readme-ov-file#usage).
149
 
150
  ---
151
 
 
117
  interface = outetts.InterfaceHF(model_version="0.3", cfg=model_config)
118
 
119
  # You can create a speaker profile for voice cloning, which is compatible across all backends.
120
+ # speaker = interface.create_speaker(audio_path="path/to/audio/file.wav")
 
 
 
 
 
121
  # interface.save_speaker(speaker, "speaker.json")
122
  # speaker = interface.load_speaker("speaker.json")
123
 
 
139
  # Save the generated speech to a file
140
  output.save("output.wav")
141
  ```
142
+ ### Additional Usage Examples
143
+
144
+ > [!IMPORTANT]
145
+ > For additional usage examples and recommendations, visit the: [GitHub repository](https://github.com/edwko/OuteTTS?tab=readme-ov-file#usage).
146
+
147
+ ### Generation Performance
148
+
149
  > [!IMPORTANT]
150
+ > The model performs best with 30-second generation batches. This window is reduced based on the length of your speaker samples. For example, if the speaker reference sample is 10 seconds, the effective window becomes approximately 20 seconds. I am currently working on adding batched generation capabilities to the library, along with further improvements that are not yet implemented.
151
 
152
  ---
153