Create app.py

app.py

@@ -0,0 +1,102 @@
+import gradio as gr
+import librosa
+import numpy as np
+import torch
+from transformers import pipeline
+
+
+language_classes = {
+    0: "Arabic",
+    1: "Basque",
+    2: "Breton",
+    3: "Catalan",
+    4: "Chinese_China",
+    5: "Chinese_Hongkong",
+    6: "Chinese_Taiwan",
+    7: "Chuvash",
+    8: "Czech",
+    9: "Dhivehi",
+    10: "Dutch",
+    11: "English",
+    12: "Esperanto",
+    13: "Estonian",
+    14: "French",
+    15: "Frisian",
+    16: "Georgian",
+    17: "German",
+    18: "Greek",
+    19: "Hakha_Chin",
+    20: "Indonesian",
+    21: "Interlingua",
+    22: "Italian",
+    23: "Japanese",
+    24: "Kabyle",
+    25: "Kinyarwanda",
+    26: "Kyrgyz",
+    27: "Latvian",
+    28: "Maltese",
+    29: "Mongolian",
+    30: "Persian",
+    31: "Polish",
+    32: "Portuguese",
+    33: "Romanian",
+    34: "Romansh_Sursilvan",
+    35: "Russian",
+    36: "Sakha",
+    37: "Slovenian",
+    38: "Spanish",
+    39: "Swedish",
+    40: "Tamil",
+    41: "Tatar",
+    42: "Turkish",
+    43: "Ukranian",
+    44: "Welsh"
+}
+
+
+
+
+username = "jpbello"  ## Complete your username
+model_id = "jpbello/Hubert_emotion-finetuned-common_language"
+device = "cuda:0" if torch.cuda.is_available() else "cpu"
+pipe = pipeline("audio-classification", model=model_id, device=device)
+
+# def predict_trunc(filepath):
+#     preprocessed = pipe.preprocess(filepath)
+#     truncated = pipe.feature_extractor.pad(preprocessed,truncation=True, max_length = 16_000*30)
+#     model_outputs = pipe.forward(truncated)
+#     outputs = pipe.postprocess(model_outputs)
+
+#     return outputs
+
+
+def classify_audio(filepath):
+    
+    
+    preds = pipe(filepath)
+    # preds = predict_trunc(filepath)
+    outputs = {}
+    for p in preds:
+        outputs[p["label"]] = p["score"]
+    return outputs
+
+
+title = "Language Classification Model"
+description = (
+    "This model has been fine-tuned on a dataset containing various languages\n"
+    "including Arabic, Basque, Catalan, Chinese, English, French, German, Japanese, Russian, and more.\n"
+    "It is designed for audio classification, allowing it to predict the language spoken in a given audio clip.\n"
+    "Try it out by uploading an audio sample and see how accurately it can identify the language being spoken!\n"
+    "For more info, check out [GITHUB](https://github.com/AEscF)"
+)
+filenames = ['EN_0049.wav', "FR_0098.wav", "JP_0222.wav",]
+filenames = [[f"./{f}"] for f in filenames]
+demo = gr.Interface(
+    fn=classify_audio,
+    inputs=gr.Audio(type="filepath"),
+    outputs=[gr.Label(label="Predictions")],
+    title=title,
+    description=description,
+    examples=filenames,
+)
+demo.launch()