Create app.py

app.py

@@ -0,0 +1,74 @@
+import torch
+
+from transformers import WhisperForConditionalGeneration, WhisperProcessor
+from transformers.models.whisper.tokenization_whisper import LANGUAGES
+from transformers.pipelines.audio_utils import ffmpeg_read
+
+import librosa
+import gradio as gr
+
+
+model_id = "openai/whisper-large-v2"
+
+processor = WhisperProcessor.from_pretrained(model_id)
+model = WhisperForConditionalGeneration.from_pretrained(model_id)
+
+sampling_rate = processor.feature_extractor.sampling_rate
+
+bos_token_id = processor.tokenizer.all_special_ids[-106]
+decoder_input_ids = torch.tensor([bos_token_id])
+
+
+def process_audio_file(file):
+    with open(file, "rb") as f:
+        inputs = f.read()
+
+    audio = ffmpeg_read(inputs, sampling_rate)
+    return audio
+
+
+def transcribe(Microphone, File_Upload):
+    warn_output = ""
+    if (Microphone is not None) and (File_Upload is not None):
+        warn_output = "WARNING: You've uploaded an audio file and used the microphone. " \
+                      "The recorded file from the microphone will be used and the uploaded audio will be discarded.\n"
+        file = Microphone
+
+    elif (Microphone is None) and (File_Upload is None):
+        return "ERROR: You have to either use the microphone or upload an audio file"
+
+    elif Microphone is not None:
+        file = Microphone
+    else:
+        file = File_Upload
+
+    audio_data = process_audio_file(file)
+
+    input_features = processor(sample["array"], sampling_rate=sample["sampling_rate"], return_tensors="pt").input_features
+    
+    with torch.no_grad():
+        logits = model.forward(input_features, decoder_input_ids=decoder_input_ids).logits
+    
+    pred_ids = torch.argmax(logits, dim=-1)
+    lang_ids = processor.decode(pred_ids[0])
+    
+    lang_ids = lang_ids.lstrip("<|").rstrip("|>")
+    language = LANGUAGES[lang_ids]
+
+    return language
+
+
+iface = gr.Interface(
+    fn=transcribe,
+    inputs=[
+        gr.inputs.Audio(source="microphone", type='filepath', optional=True),
+        gr.inputs.Audio(source="upload", type='filepath', optional=True),
+    ],
+    outputs="text",
+    layout="horizontal",
+    theme="huggingface",
+    title="Whisper Language Identification",
+    description="Demo for Language Identification using OpenAI's [Whisper Large V2](https://huggingface.co/openai/whisper-large-v2)",
+    allow_flagging='never',
+)
+iface.launch(enable_queue=True)