Update app.py
Browse files
app.py
CHANGED
@@ -23,6 +23,8 @@ asr_model.change_decoding_strategy(None)
|
|
23 |
decoding_cfg = asr_model.cfg.decoding
|
24 |
decoding_cfg.beam.beam_size = 1
|
25 |
asr_model.change_decoding_strategy(decoding_cfg)
|
|
|
|
|
26 |
feature_stride = asr_model.cfg.preprocessor['window_stride']
|
27 |
model_stride_in_secs = feature_stride * 8
|
28 |
frame_asr = FrameBatchMultiTaskAED(
|
@@ -130,6 +132,6 @@ gr.Interface(
|
|
130 |
gr.Textbox(label="Generated Text"),
|
131 |
gr.Audio(type="filepath", label="Generated Speech")
|
132 |
],
|
133 |
-
title="
|
134 |
-
description="
|
135 |
).launch(inbrowser=True)
|
|
|
23 |
decoding_cfg = asr_model.cfg.decoding
|
24 |
decoding_cfg.beam.beam_size = 1
|
25 |
asr_model.change_decoding_strategy(decoding_cfg)
|
26 |
+
asr_model.cfg.preprocessor.dither = 0.0
|
27 |
+
asr_model.cfg.preprocessor.pad_to = 0
|
28 |
feature_stride = asr_model.cfg.preprocessor['window_stride']
|
29 |
model_stride_in_secs = feature_stride * 8
|
30 |
frame_asr = FrameBatchMultiTaskAED(
|
|
|
132 |
gr.Textbox(label="Generated Text"),
|
133 |
gr.Audio(type="filepath", label="Generated Speech")
|
134 |
],
|
135 |
+
title="YOUR AWESOME AI ASSISTANT",
|
136 |
+
description="Gets input audio from user, transcribe it with ASR Canary1b, generate text with Phi3LLM, and convert it back to speech with VITS TTS."
|
137 |
).launch(inbrowser=True)
|