feat: gradio app

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,4 @@
+__pycache__
+*.pyc
+.gradio
+.DS_Store

app.py

@@ -0,0 +1,49 @@
+import gradio as gr
+from huggingface_hub import InferenceClient
+
+model_id = "JacobLinCool/whisper-large-v3-turbo-common_voice_16_1-zh-TW-2"
+client = InferenceClient(model_id)
+
+
+def transcribe_audio(audio: str) -> str:
+    out = client.automatic_speech_recognition(audio)
+    return out.text
+
+
+with gr.Blocks() as demo:
+    gr.Markdown("# TWASR: Chinese (Taiwan) Automatic Speech Recognition.")
+    gr.Markdown("Upload an audio file or record your voice to transcribe it to text.")
+    gr.Markdown(
+        "First load may take a while to initialize the model, following requests will be faster."
+    )
+
+    with gr.Row():
+        audio_input = gr.Audio(
+            label="Audio", type="filepath", show_download_button=True
+        )
+        text_output = gr.Textbox(label="Transcription")
+
+    transcribe_button = gr.Button("Transcribe with Inference API")
+    transcribe_button.click(
+        fn=transcribe_audio, inputs=[audio_input], outputs=[text_output]
+    )
+
+    gr.Examples(
+        [
+            ["./examples/audio1.mp3"],
+            ["./examples/audio2.mp3"],
+        ],
+        inputs=[audio_input],
+        outputs=[text_output],
+        fn=transcribe_audio,
+        cache_examples=True,
+        cache_mode="lazy",
+        run_on_click=True,
+    )
+
+    gr.Markdown(
+        f"Current model: {model_id}. For more information, visit the [model hub](https://huggingface.co/{model_id})."
+    )
+
+if __name__ == "__main__":
+    demo.launch()

examples/audio1.mp3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:033a08d5a16f5f6f6ddfa7dd6eea87c15f06060bc08ebbb2b2746579022bac13
+size 40557

examples/audio2.mp3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d2cca183511d5db331624d01f4f3a01404052eaba3bbfa6c0c2f2bcd50b8934
+size 33837

requirements.txt

@@ -0,0 +1,2 @@
+gradio==5.4.0
+huggingface_hub==0.26.2