Upload 5 files

.gitattributes

@@ -33,3 +33,4 @@ 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
+stt_kz_quartznet15x5.nemo filter=lfs diff=lfs merge=lfs -text

app.py

@@ -0,0 +1,16 @@
+import os
+import gradio as gr
+from model.model import transcribe
+
+
+title = "Automatic Speech Recognition Using NVIDIA NeMo for Kazakh Speech"
+example_list = [["examples/" + example] for example in os.listdir("examples")]
+
+demo = gr.Interface(
+    fn=transcribe, 
+    inputs=gr.Audio(source="microphone", type="filepath"), 
+    outputs="text",
+    title=title,
+    examples=example_list)
+
+demo.launch()

stt_kz_quartznet15x5.nemo

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f5dd5ae4ebef4832e19ed80d70ecfd1979e8a1d2158474cc2a986f365b41501
+size 76380160

transcribe.py

@@ -0,0 +1,31 @@
+import os
+from pathlib import Path
+import nemo.collections.asr as nemo_asr
+
+
+def converter(audio_file):
+    converted = audio_file.split(".")[0] + "converted_.wav"
+    cmd_str = f"ffmpeg -y -i {audio_file} -ac 1 -ar 16000 {converted}"
+    os.system(cmd_str)
+    # os.remove(audio_file)
+    return converted
+    
+    
+
+def transcribe(audio_file):
+    wav_file = converter(audio_file)
+    try:
+        text = model_kz.transcribe([wav_file])
+        return text[0]
+    except:
+        return 'Try another file format.'
+    
+
+
+
+language = "kz"
+
+BASE_DIR = Path(__file__).resolve(strict=True).parent
+
+model_kz = nemo_asr.models.EncDecCTCModel.restore_from(restore_path=f"{BASE_DIR}/stt_{language}_quartznet15x5.nemo")
+