feat: Improve description

app.py

@@ -12,11 +12,22 @@ from transformers import pipeline
 
 warnings.filterwarnings("ignore", category=FutureWarning)
 
+icon = """
+<svg xmlns="http://www.w3.org/2000/svg" width="14px" viewBox="0 0 24 24" fill="none"
+     stroke="currentColor" stroke-width="2" stroke-linecap="round"
+     stroke-linejoin="round" style="display: inline;">
+  <path d="M21 15v4a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2v-4"/>
+  <polyline points="17 8 12 3 7 8"/>
+  <line x1="12" y1="3" x2="12" y2="15"/>
+</svg>
+"""
 TITLE = "Røst ASR Demo"
-DESCRIPTION = """
+DESCRIPTION = f"""
 This is a demo of the Danish speech recognition model
-[Røst](https://huggingface.co/alexandrainst/roest-315m). Speak into the microphone and
-see the text appear on the screen!
+[Røst](https://huggingface.co/alexandrainst/roest-315m). Press "Record" to record your
+own voice. When you're done you can press "Stop" to stop recording and "Submit" to
+send the audio to the model for transcription. You can also upload an audio file by
+pressing the {icon} button.
 """
 
 device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
@@ -24,7 +35,7 @@ transcriber = pipeline(
     task="automatic-speech-recognition",
     model="alexandrainst/roest-315m",
     device=device,
-    token=os.environ["HUGGINGFACE_HUB_TOKEN"],
+    token=os.getenv("HUGGINGFACE_HUB_TOKEN", True),
 )
 transcription_fixer = PunctFixer(language="da", device=device)