Upload 4 files

.gitattributes

@@ -32,3 +32,4 @@ 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
+audio/audio.mp3 filter=lfs diff=lfs merge=lfs -text

app.py

@@ -0,0 +1,85 @@
+import gradio as gr
+import os 
+
+from helpers import make_header, upload_file, request_transcript, wait_for_completion, make_paragraphs_string
+
+
+title = """<h1 align="center">🔥AssemblyAI: Conformer-1 Demo🔥</h1>"""
+
+subtitle = """<h2 align="center">Automatic Speech Recognition using the AssemblyAI API</h2>"""
+link = """<p align="center"><a href="https://www.assemblyai.com/blog/conformer-1/">Click here to learn more about the Conformer-1 model</a></p>"""
+
+
+def submit_to_AAI(api_key,
+                  radio,
+                  audio_file,
+                  mic_recording):
+
+    if radio == "Audio File":
+        audio_data = audio_file
+    elif radio == "Record Audio":
+        audio_data = mic_recording
+        
+    header = make_header(api_key)
+        
+    # 1. Upload the audio
+    upload_url = upload_file(audio_data, header, is_file=False)
+
+    # 2. Request transcript
+    transcript_response = request_transcript(upload_url, header)
+    
+    transcript_id = transcript_response['id']
+    
+    # 3. Wait for the transcription to complete
+    _, error = wait_for_completion(transcript_id, header)
+    
+    if error is not None:
+        return error
+
+    # 4. Fetch paragraphs of transcript    
+    return make_paragraphs_string(transcript_id, header)
+
+
+def change_audio_source(radio):
+    if radio == "Audio File":
+        return [gr.Audio.update(visible=True),
+                gr.Audio.update(visible=False)]
+    elif radio == "Record Audio":
+        return [gr.Audio.update(visible=False),
+                gr.Audio.update(visible=True)]
+                
+with gr.Blocks(css = """#col_container {width: 1000px; margin-left: auto; margin-right: auto;}
+                #chatbot {height: 520px; overflow: auto;}""") as demo:
+    gr.HTML('<center><a href="https://www.assemblyai.com/"><img src="file/images/logo.png" width="180px"></a></center>')
+    gr.HTML(title)
+    gr.HTML(subtitle)
+    gr.HTML(link)
+    gr.HTML('''<center><a href="https://huggingface.co/spaces/assemblyai/Conformer1-Demo?duplicate=true"><img src="https://bit.ly/3gLdBN6" alt="Duplicate Space"></a>Duplicate the Space and run securely with your AssemblyAI API Key</center>''')
+       
+    with gr.Column(elem_id="col_container"):
+        api_key = gr.Textbox(type='password', label="Enter your AssemblyAI API key here")
+        
+        with gr.Box():
+            # Selector for audio source
+            radio = gr.Radio(["Audio File", "Record Audio"], label="Audio Source", value="Audio File")    
+            # Audio object for both file and microphone data
+            audio_file = gr.Audio()
+            mic_recording = gr.Audio(source="microphone", visible=False)
+        
+            gr.Examples([os.path.join(os.path.dirname(__file__),"audio/audio.mp3")], audio_file)
+        
+        btn = gr.Button("Run")
+        
+        out = gr.Textbox(placeholder="Your formatted transcript will appear here ...", lines=10)
+        
+        # Changing audio source changes Audio input component
+        radio.change(fn=change_audio_source,
+                    inputs=[radio],
+                    outputs=[audio_file, mic_recording])
+            
+        # Clicking "submit" uploads selected audio to AssemblyAI, performs requested analyses, and displays results
+        btn.click(fn=submit_to_AAI,
+                inputs=[api_key,radio,audio_file,mic_recording],
+                outputs=out)
+
+    demo.launch(debug=True)

audio/audio.mp3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37d851f5525c4b54b3c565f46fa47105f5c9533deed15eb7e6874f31b340659b
+size 2353876

helpers.py

@@ -0,0 +1,98 @@
+import requests
+import time
+from scipy.io.wavfile import write
+import io
+
+
+upload_endpoint = "https://api.assemblyai.com/v2/upload"
+transcript_endpoint = "https://api.assemblyai.com/v2/transcript"
+
+
+def make_header(api_key):
+    return {
+        'authorization': api_key,
+        'content-type': 'application/json'
+    }
+
+
+def _read_file(filename, chunk_size=5242880):
+    """Reads the file in chunks. Helper for `upload_file()`"""
+    with open(filename, "rb") as f:
+        while True:
+            data = f.read(chunk_size)
+            if not data:
+                break
+            yield data
+
+
+def _read_array(audio, chunk_size=5242880):
+    """Like _read_file but for array - creates temporary unsaved "file" from sample rate and audio np.array"""
+    sr, aud = audio
+
+    # Create temporary "file" and write data to it
+    bytes_wav = bytes()
+    temp_file = io.BytesIO(bytes_wav)
+    write(temp_file, sr, aud)
+
+    while True:
+        data = temp_file.read(chunk_size)
+        if not data:
+            break
+        yield data
+
+
+def upload_file(audio_file, header, is_file=True):
+    """Uploads a file to AssemblyAI"""
+    upload_response = requests.post(
+        upload_endpoint,
+        headers=header,
+        data=_read_file(audio_file) if is_file else _read_array(audio_file)
+    )
+    if upload_response.status_code != 200:
+        upload_response.raise_for_status()
+    # Returns {'upload_url': <URL>}
+    return upload_response.json()
+
+
+def request_transcript(upload_url, header):
+    """Requests a transcript from AssemblyAI"""
+
+    # If input is a dict returned from `upload_file` rather than a raw upload_url string
+    if type(upload_url) is dict:
+        upload_url = upload_url['upload_url']
+
+    # Create request
+    transcript_request = {
+        'audio_url': upload_url,
+    }
+
+    # POST request
+    transcript_response = requests.post(
+        transcript_endpoint,
+        json=transcript_request,
+        headers=header
+    )
+
+    return transcript_response.json()
+
+
+def wait_for_completion(transcript_id, header):
+    """Given a polling endpoint, waits for the transcription/audio analysis to complete"""
+    polling_endpoint = "https://api.assemblyai.com/v2/transcript/" + transcript_id
+    
+    while True:
+        polling_response = requests.get(polling_endpoint, headers=header)
+        polling_response = polling_response.json()
+
+        if polling_response['status'] == 'completed':
+            return polling_response, None
+        elif polling_response['status'] == 'error':
+            return None, f"Error: {polling_response['error']}"
+
+        time.sleep(5)
+
+
+def make_paragraphs_string(transc_id, header):
+    endpoint = transcript_endpoint + "/" + transc_id + "/paragraphs"
+    paras = requests.get(endpoint, headers=header).json()['paragraphs']
+    return '\n\n'.join(i['text'] for i in paras)