Skip to content

SOTA Open Source TTS

update webui

Lengyue1e7a2b0a16c4
diff --git a/fish_speech/text/clean.py b/fish_speech/text/clean.py
index 76d9dc9..3f7eb0e 100644
--- a/fish_speech/text/clean.py
+++ b/fish_speech/text/clean.py
@@ -64,6 +64,6 @@ def clean_text(text):
# Replace all chinese symbols with their english counterparts
text = REPLACE_SYMBOL_REGEX.sub(lambda x: SYMBOLS_MAPPING[x.group()], text)
- text = REMOVE_UNKNOWN_SYMBOL_REGEX.sub("", text)
+ # text = REMOVE_UNKNOWN_SYMBOL_REGEX.sub("", text)
return text
diff --git a/tools/api.py b/tools/api.py
index e5331a2..b15c89a 100644
--- a/tools/api.py
+++ b/tools/api.py
@@ -148,7 +148,7 @@ def decode_vq_tokens(
return decoder_model.decode(
indices=codes[None],
feature_lengths=feature_lengths,
- ).squeeze()
+ )[0].squeeze()
raise ValueError(f"Unknown model type: {type(decoder_model)}")
diff --git a/tools/webui.py b/tools/webui.py
index 9140262..6271ff1 100644
--- a/tools/webui.py
+++ b/tools/webui.py
@@ -40,7 +40,7 @@ HEADER_MD = f"""# Fish Speech
{i18n("A text-to-speech model based on VQ-GAN and Llama developed by [Fish Audio](https://fish.audio).")}
-{i18n("You can find the source code [here](https://github.com/fishaudio/fish-speech) and models [here](https://huggingface.co/fishaudio/fish-speech-1).")}
+{i18n("You can find the source code [here](https://github.com/fishaudio/fish-speech) and models [here](https://huggingface.co/fishaudio/fish-speech-1.4).")}
{i18n("Related code and weights are released under CC BY-NC-SA 4.0 License.")}
@@ -367,7 +367,7 @@ def build_app():
with gr.Row():
if_refine_text = gr.Checkbox(
label=i18n("Text Normalization"),
- value=True,
+ value=False,
scale=1,
)
@@ -381,9 +381,9 @@ def build_app():
with gr.Tab(label=i18n("Advanced Config")):
chunk_length = gr.Slider(
label=i18n("Iterative Prompt Length, 0 means off"),
- minimum=0,
- maximum=500,
- value=100,
+ minimum=50,
+ maximum=300,
+ value=200,
step=8,
)