Add language selection

app.py

@@ -7,17 +7,19 @@ from share_btn import community_icon_html, loading_icon_html, share_js
 
 model = whisper.load_model("medium")
 
+languages = {long_name: short_name for short_name, long_name in whisper.tokenizer.LANGUAGES.items()}
+
 
         
-def inference(audio):
+def inference(audio, language):
     audio = whisper.load_audio(audio)
     audio = whisper.pad_or_trim(audio)
     
     mel = whisper.log_mel_spectrogram(audio).to(model.device)
     
-    _, probs = model.detect_language(mel)
+    #_, probs = model.detect_language(mel)
     
-    options = whisper.DecodingOptions(fp16 = False)
+    options = whisper.DecodingOptions(fp16 = False, language=language)
     result = whisper.decode(model, mel, options)
     
     print(result.text)
@@ -146,7 +148,11 @@ with block:
                     type="filepath"
                 )
 
-                btn = gr.Button("Transcribe")
+                language_long_name = gr.Dropdown(
+                    list(languages.keys()), value="auto", multiselect=False, label="Taal", info="Taal van de gesproken tekst. Auto voor automatische detectie."
+                )
+
+                btn = gr.Button("Transcribeer")
         text = gr.Textbox(show_label=False, elem_id="result-textarea")
         with gr.Group(elem_id="share-btn-container"):
             community_icon = gr.HTML(community_icon_html, visible=False)
@@ -156,7 +162,7 @@ with block:
 
 
         
-        btn.click(inference, inputs=[audio], outputs=[text, community_icon, loading_icon, share_button])
+        btn.click(inference, inputs=[audio, languages[language_long_name]], outputs=[text, community_icon, loading_icon, share_button])
         share_button.click(None, [], [], _js=share_js)
  
         gr.HTML('''