From a3e1fa35187b142f43980a58d9ae4eb4f785f448 Mon Sep 17 00:00:00 2001 From: mrq Date: Sun, 17 Nov 2024 09:28:33 -0600 Subject: [PATCH] ugh --- vall_e/models/base.py | 2 ++ vall_e/webui.py | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/vall_e/models/base.py b/vall_e/models/base.py index 3c619d1..d043b2d 100755 --- a/vall_e/models/base.py +++ b/vall_e/models/base.py @@ -1221,8 +1221,10 @@ class Base(nn.Module): ) """ + """ if classifier_level == "AR:0:0": classifier_level = "NAR:0:0" + """ embedding = self.resps_emb( input if input.dim() == 1 or quant_level == 0 else input[:, :quant_level], diff --git a/vall_e/webui.py b/vall_e/webui.py index 41e8c3a..34d9272 100644 --- a/vall_e/webui.py +++ b/vall_e/webui.py @@ -464,7 +464,7 @@ with ui: with gr.Row(): layout["inference_tts"]["inputs"]["input-prompt-prefix"] = gr.Checkbox(label="Input Prompt as Prefix", info="Treats the input prompt clip as the prefix of the generated sequence.") layout["inference_tts"]["inputs"]["prefix-silence"] = gr.Slider(value=0.0, minimum=0.0, maximum=1.0, step=0.05, label="Silence Prefix Duration", info="Amount of silence to prefix to the output response before beginning inference.") - layout["inference_tts"]["inputs"]["modality"] = gr.Dropdown(value="AR+NAR", choices=["AR+NAR", "NAR-len"], label="Modality", info="Whether to inference with the AR+NAR or through the NAR-len.") + layout["inference_tts"]["inputs"]["modality"] = gr.Dropdown(value="Auto", choices=["Auto", "AR+NAR", "NAR-len"], label="Modality", info="Whether to inference with the AR+NAR or through the NAR-len.") with gr.Row(): layout["inference_tts"]["inputs"]["beam-width"] = gr.Slider(value=0, minimum=0, maximum=32, step=1, label="Beam Width", info="Number of branches to search through for beam search sampling.") layout["inference_tts"]["inputs"]["dynamic-sampling"] = gr.Checkbox(label="Dynamic Temperature", info="Dynamically adjusts the temperature based on the highest confident predicted token per sampling step.")