The app

README.md

@@ -10,4 +10,4 @@ pinned: false
 license: cc-by-nc-sa-4.0
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces#reference
+Czech Wav2Vec2-XLS-R model from 250h of recordings.

app.py

@@ -0,0 +1,29 @@
+from transformers import pipeline, AutoFeatureExtractor, AutoTokenizer, Wav2Vec2ForCTC
+import gradio as gr
+import time
+
+model_id = 'comodoro/wav2vec2-xls-r-300m-cs-250'
+feature_extractor = AutoFeatureExtractor.from_pretrained(model_id)
+model = Wav2Vec2ForCTC.from_pretrained(model_id)
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+
+p = pipeline("automatic-speech-recognition", chunk_length_s=5, model=model, 
+    tokenizer=tokenizer, feature_extractor=feature_extractor)
+
+def transcribe(audio, state=""):
+    time.sleep(2)
+    text = p(audio)["text"]
+    state += text + " "
+    return state, state
+
+gr.Interface(
+    fn=transcribe, 
+    inputs=[
+        gr.inputs.Audio(source="microphone", type="filepath"), 
+        "state"
+    ],
+    outputs=[
+        "textbox",
+        "state"
+    ],
+    live=True).launch()

packages.txt

@@ -0,0 +1 @@
+ffmpeg

requirements.txt

@@ -0,0 +1,2 @@
+torch
+transformers