Update app.py

app.py

@@ -134,69 +134,8 @@ scheduler = CommitScheduler(
     allow_patterns=DB_NAME,
 )
 
-# Load audio dataset
-# audio_dataset = load_dataset(AUDIO_DATASET_ID)
 
-####################################
-# Router API
-####################################
-router = Client("TTS-AGI/tts-router", hf_token=os.getenv('HF_TOKEN'))
-####################################
-# Gradio app
-####################################
-MUST_BE_LOGGEDIN = "Please login with Hugging Face to participate in the TTS Arena."
-DESCR = """
-# TTS Arena: Benchmarking TTS Models in the Wild
-Vote to help the community find the best available text-to-speech model!
-""".strip()
-# INSTR = """
-# ## Instructions
-
-# * Listen to two anonymous models
-# * Vote on which synthesized audio sounds more natural to you
-# * If there's a tie, click Skip
-
-# **When you're ready to begin, login and begin voting!** The model names will be revealed once you vote.
-# """.strip()
-INSTR = """
-## 🗳️ Vote
-* Input text (English only) to synthesize audio (or press 🎲 for random text).
-* Listen to the two audio clips, one after the other.
-* Vote on which audio sounds more natural to you.
-* _Note: Model names are revealed after the vote is cast._
-Note: It may take up to 30 seconds to synthesize audio.
-""".strip()
-request = ''
-if SPACE_ID:
-    request = f"""
-### Request a model
-Please [create a Discussion](https://huggingface.co/spaces/{SPACE_ID}/discussions/new) to request a model.
-"""
-ABOUT = f"""
-## 📄 About
-The TTS Arena evaluates leading speech synthesis models. It is inspired by LMsys's [Chatbot Arena](https://chat.lmsys.org/).
-### Motivation
-The field of speech synthesis has long lacked an accurate method to measure the quality of different models. Objective metrics like WER (word error rate) are unreliable measures of model quality, and subjective measures such as MOS (mean opinion score) are typically small-scale experiments conducted with few listeners. As a result, these measurements are generally not useful for comparing two models of roughly similar quality. To address these drawbacks, we are inviting the community to rank models in an easy-to-use interface, and opening it up to the public in order to make both the opportunity to rank models, as well as the results, more easily accessible to everyone.
-### The Arena
-The leaderboard allows a user to enter text, which will be synthesized by two models. After listening to each sample, the user can vote on which model sounds more natural. Due to the risks of human bias and abuse, model names are revealed only after a vote is submitted.
-### Credits
-Thank you to the following individuals who helped make this project possible:
-* VB ([Twitter](https://twitter.com/reach_vb) / [Hugging Face](https://huggingface.co/reach-vb))
-* Clémentine Fourrier ([Twitter](https://twitter.com/clefourrier) / [Hugging Face](https://huggingface.co/clefourrier))
-* Lucain Pouget ([Twitter](https://twitter.com/Wauplin) / [Hugging Face](https://huggingface.co/Wauplin))
-* Yoach Lacombe ([Twitter](https://twitter.com/yoachlacombe) / [Hugging Face](https://huggingface.co/ylacombe))
-* Main Horse ([Twitter](https://twitter.com/main_horse) / [Hugging Face](https://huggingface.co/main-horse))
-* Sanchit Gandhi ([Twitter](https://twitter.com/sanchitgandhi99) / [Hugging Face](https://huggingface.co/sanchit-gandhi))
-* Apolinário Passos ([Twitter](https://twitter.com/multimodalart) / [Hugging Face](https://huggingface.co/multimodalart))
-* Pedro Cuenca ([Twitter](https://twitter.com/pcuenq) / [Hugging Face](https://huggingface.co/pcuenq))
-{request}
-### Privacy statement
-We may store text you enter and generated audio. We store a unique ID for each session. You agree that we may collect, share, and/or publish any data you input for research and/or commercial purposes.
-### License
-Generated audio clips cannot be redistributed and may be used for personal, non-commercial use only.
-Random sentences are sourced from a filtered subset of the [Harvard Sentences](https://www.cs.columbia.edu/~hgs/audio/harvard.html).
-""".strip()
-LDESC = """
+
 ## 🏆 Leaderboard
 Vote to help the community determine the best text-to-speech (TTS) models.
 The leaderboard displays models in descending order of how natural they sound (based on votes cast by the community).
@@ -205,12 +144,6 @@ Important: In order to help keep results fair, the leaderboard hides results by
 
 
 
-
-# def reload_audio_dataset():
-#     global audio_dataset
-#     audio_dataset = load_dataset(AUDIO_DATASET_ID)
-#     return 'Reload Audio Dataset'
-
 def del_db(txt):
     if not txt.lower() == 'delete db':
         raise gr.Error('You did not enter "delete db"')
@@ -298,17 +231,7 @@ model_links = {
     'speecht5': 'https://github.com/microsoft/SpeechT5',
     'metavoice': 'https://github.com/metavoiceio/metavoice-src',
 }
-# def get_random_split(existing_split=None):
-#     choice = random.choice(list(audio_dataset.keys()))
-#     if existing_split and choice == existing_split:
-#         return get_random_split(choice)
-#     else:
-#         return choice
-
-# def get_random_splits():
-#     choice1 = get_random_split()
-#     choice2 = get_random_split(choice1)
-#     return (choice1, choice2)
+
 def model_license(name):
     print(name)
     for k, v in AVAILABLE_MODELS.items():
@@ -317,6 +240,8 @@ def model_license(name):
                 return model_licenses[v]
     print('---')
     return 'Unknown'
+
+
 def get_leaderboard(reveal_prelim = False):
     conn = get_db()
     cursor = conn.cursor()
@@ -350,10 +275,13 @@ def get_leaderboard(reveal_prelim = False):
     # df = df[['order', 'name', 'score', 'license', 'votes']]
     df = df[['order', 'name', 'score', 'votes']]
     return df
+
+
 def mkuuid(uid):
     if not uid:
         uid = uuid.uuid4()
     return uid
+
 def upvote_model(model, uname):
     conn = get_db()
     cursor = conn.cursor()
@@ -364,6 +292,8 @@ def upvote_model(model, uname):
     with scheduler.lock:
         conn.commit()
     cursor.close()
+
+
 def log_text(text):
     conn = get_db()
     cursor = conn.cursor()
@@ -371,6 +301,8 @@ def log_text(text):
     with scheduler.lock:
         conn.commit()
     cursor.close()
+
+
 def downvote_model(model, uname):
     conn = get_db()
     cursor = conn.cursor()
@@ -471,51 +403,7 @@ with gr.Blocks() as leaderboard:
     reloadbtn.click(get_leaderboard, inputs=[reveal_prelim], outputs=[df])
     # gr.Markdown("DISCLAIMER: The licenses listed may not be accurate or up to date, you are responsible for checking the licenses before using the models. Also note that some models may have additional usage restrictions.")
 
-# with gr.Blocks() as vote:
-#     useridstate = gr.State()
-#     gr.Markdown(INSTR)
-#     # gr.LoginButton()
-#     with gr.Row():
-#         gr.HTML('<div align="left"><h3>Model A</h3></div>')
-#         gr.HTML('<div align="right"><h3>Model B</h3></div>')
-#     model1 = gr.Textbox(interactive=False, visible=False, lines=1, max_lines=1)
-#     model2 = gr.Textbox(interactive=False, visible=False, lines=1, max_lines=1)
-#     # with gr.Group():
-#     #     with gr.Row():
-#     #         prevmodel1 = gr.Textbox(interactive=False, show_label=False, container=False, value="Vote to reveal model A")
-#     #         prevmodel2 = gr.Textbox(interactive=False, show_label=False, container=False, value="Vote to reveal model B", text_align="right")
-#     #     with gr.Row():
-#     #         aud1 = gr.Audio(interactive=False, show_label=False, show_download_button=False, show_share_button=False, waveform_options={'waveform_progress_color': '#3C82F6'})
-#     #         aud2 = gr.Audio(interactive=False, show_label=False, show_download_button=False, show_share_button=False, waveform_options={'waveform_progress_color': '#3C82F6'})
-#     with gr.Group():
-#         with gr.Row():
-#             with gr.Column():
-#                 with gr.Group():
-#                     prevmodel1 = gr.Textbox(interactive=False, show_label=False, container=False, value="Vote to reveal model A", lines=1, max_lines=1)
-#                     aud1 = gr.Audio(interactive=False, show_label=False, show_download_button=False, show_share_button=False, waveform_options={'waveform_progress_color': '#3C82F6'})
-#             with gr.Column():
-#                 with gr.Group():
-#                     prevmodel2 = gr.Textbox(interactive=False, show_label=False, container=False, value="Vote to reveal model B", text_align="right", lines=1, max_lines=1)
-#                     aud2 = gr.Audio(interactive=False, show_label=False, show_download_button=False, show_share_button=False, waveform_options={'waveform_progress_color': '#3C82F6'})
-
-
-#     with gr.Row():
-#         abetter = gr.Button("A is Better", variant='primary', scale=4)
-#         # skipbtn = gr.Button("Skip", scale=1)
-#         bbetter = gr.Button("B is Better", variant='primary', scale=4)
-#     with gr.Row():
-#         bothbad = gr.Button("Both are Bad", scale=2)
-#         skipbtn = gr.Button("Skip", scale=1)
-#         bothgood = gr.Button("Both are Good", scale=2)
-#     outputs = [aud1, aud2, model1, model2, useridstate, prevmodel1, prevmodel2]
-#     abetter.click(a_is_better, outputs=outputs, inputs=[model1, model2, useridstate])
-#     bbetter.click(b_is_better, outputs=outputs, inputs=[model1, model2, useridstate])
-#     skipbtn.click(b_is_better, outputs=outputs, inputs=[model1, model2, useridstate])
-
-#     bothbad.click(both_bad, outputs=outputs, inputs=[model1, model2, useridstate])
-#     bothgood.click(both_good, outputs=outputs, inputs=[model1, model2, useridstate])
-
-#     vote.load(reload, outputs=[aud1, aud2, model1, model2])
+
 def doloudnorm(path):
     data, rate = sf.read(path)
     meter = pyln.Meter(rate)