Inspired by the post of @gaborfodor, I have updated my streamlit existing app and added his PANN predictive model and some plots from his H2o wave app.
\nYou can find the app live here and here is a screenshot:
\nLet me know what you think and which features you would like to see.
\nSome links:
\n", "messages": [ { "id": "1337580", "postDate": "06/05/2021 17:54:22", "content": "Inspired by the post of @gaborfodor, I have updated my streamlit existing app and added his PANN predictive model and some plots from his H2o wave app.
\nYou can find the app live here and here is a screenshot:
\nLet me know what you think and which features you would like to see.
\nSome links:
\n", "rawMarkdown": "Inspired by the [post](https://www.kaggle.com/c/birdclef-2021/discussion/243390#1337160) of @gaborfodor, I have updated my **[streamlit](https://streamlit.io/)** existing app and added his **PANN predictive model** and some plots from his H2o wave app. \n\nYou can find the app live [here](https://share.streamlit.io/yassinealouini/rfcx-species-audio-detection-streamlit/app.py) and here is a screenshot: \n\n\n\n\n\nLet me know what you think and which features you would like to see.\n\n\nSome links:\n\n- The original app source code [repo](https://github.com/gaborfodor/wave-bird-recognition)\n- The streamlit app source code [repo](https://github.com/yassineAlouini/rfcx-species-audio-detection-streamlit)", "votes": null }, { "id": "1337651", "postDate": "06/05/2021 18:55:12", "content": "It looks great!
\nI tried it out with a sample 60-sec clip from Rainforest audio competition, but the app crashed by the way I hear the output audio (FYI the predictions has been returned):
Oh no.\n\nError running app. If this keeps happening, please contact support.\nThanks for giving it a try. Can you share your file please so that I can try to debug it. 👌
", "rawMarkdown": "Thanks for giving it a try. Can you share your file please so that I can try to debug it. 👌", "votes": null }, { "id": "1337678", "postDate": "06/05/2021 19:18:38", "content": "@yassinealouini
\nI picked a .flac file from here:
\nhttps://www.kaggle.com/alexlwh/rfcx-small-train-data?select=train_small
I will give it a try. It might be the file format. I will post my findings. 👌
", "rawMarkdown": "I will give it a try. It might be the file format. I will post my findings. 👌", "votes": null }, { "id": "1337690", "postDate": "06/05/2021 19:33:06", "content": "Alright, I had to reboot the app since it went down.
\nThen, I tried one .flac file and it worked.
Could you retry @alexlwh please? By the way, can you share your browser version and type? Thanks for helping me investigate. :)
", "rawMarkdown": "Alright, I had to reboot the app since it went down. \n\nThen, I tried one `.flac` file and it worked. \n\n\n\n\nCould you retry @alexlwh please? By the way, can you share your browser version and type? Thanks for helping me investigate. :)", "votes": null }, { "id": "1337699", "postDate": "06/05/2021 19:50:07", "content": "One other hypothesis I have is that due to how the application is deployed, it could be rebooting each time too much traffic is reached. 😄
", "rawMarkdown": "One other hypothesis I have is that due to how the application is deployed, it could be rebooting each time too much traffic is reached. 😄", "votes": null }, { "id": "1337702", "postDate": "06/05/2021 19:53:10", "content": "So one solution is to get the app's source code and deploy it on your own. It is very easy using the streamlit deploy documentation page.
", "rawMarkdown": "So one solution is to get the app's source code and deploy it on your own. It is very easy using the streamlit deploy documentation [page](https://docs.streamlit.io/en/stable/deploy_streamlit_app.html).", "votes": null }, { "id": "1337981", "postDate": "06/06/2021 04:54:23", "content": "Perfect! it works for me now
\nIt looks pretty nice 👍
What did u use for deployment btw?
", "rawMarkdown": "What did u use for deployment btw?", "votes": null }, { "id": "1338044", "postDate": "06/06/2021 06:00:42", "content": "I am using the streamlit's own infra but you can do it with many different providers. As said above, the
\ndeploy page is very helpful.
Basically, you need:
\nIf you have trouble doing it, let me know.
", "rawMarkdown": "I am using the streamlit's own infra but you can do it with many different providers. As said above, the \ndeploy [page](https://docs.streamlit.io/en/stable/deploy_streamlit_app.html) is very helpful. \n\nBasically, you need:\n\n- sign up to streamlit (using Github is the easisest way)\n- create a Github repo (should be public to have access to the free tier)\n- deploy, you have 3 free apps\n\nIf you have trouble doing it, let me know.", "votes": null } ], "comments": [ { "id": 1337651, "author_name": "alexlwh", "author_url": "", "post_date": "06/05/2021 18:55:12", "content": "It looks great!
\nI tried it out with a sample 60-sec clip from Rainforest audio competition, but the app crashed by the way I hear the output audio (FYI the predictions has been returned):
Oh no.\n\nError running app. If this keeps happening, please contact support.\nThanks for giving it a try. Can you share your file please so that I can try to debug it. 👌
", "votes": null, "replies": [] }, { "id": 1337678, "author_name": "alexlwh", "author_url": "", "post_date": "06/05/2021 19:18:38", "content": "@yassinealouini
\nI picked a .flac file from here:
\nhttps://www.kaggle.com/alexlwh/rfcx-small-train-data?select=train_small
I will give it a try. It might be the file format. I will post my findings. 👌
", "votes": null, "replies": [] }, { "id": 1337690, "author_name": "yassinealouini", "author_url": "", "post_date": "06/05/2021 19:33:06", "content": "Alright, I had to reboot the app since it went down.
\nThen, I tried one .flac file and it worked.
Could you retry @alexlwh please? By the way, can you share your browser version and type? Thanks for helping me investigate. :)
", "votes": null, "replies": [] }, { "id": 1337699, "author_name": "yassinealouini", "author_url": "", "post_date": "06/05/2021 19:50:07", "content": "One other hypothesis I have is that due to how the application is deployed, it could be rebooting each time too much traffic is reached. 😄
", "votes": null, "replies": [] }, { "id": 1337702, "author_name": "yassinealouini", "author_url": "", "post_date": "06/05/2021 19:53:10", "content": "So one solution is to get the app's source code and deploy it on your own. It is very easy using the streamlit deploy documentation page.
", "votes": null, "replies": [] }, { "id": 1337981, "author_name": "alexlwh", "author_url": "", "post_date": "06/06/2021 04:54:23", "content": "Perfect! it works for me now
\nIt looks pretty nice 👍
What did u use for deployment btw?
", "votes": null, "replies": [] }, { "id": 1338044, "author_name": "yassinealouini", "author_url": "", "post_date": "06/06/2021 06:00:42", "content": "I am using the streamlit's own infra but you can do it with many different providers. As said above, the
\ndeploy page is very helpful.
Basically, you need:
\nIf you have trouble doing it, let me know.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1337580": "Inspired by the [post](https://www.kaggle.com/c/birdclef-2021/discussion/243390#1337160) of @gaborfodor, I have updated my **[streamlit](https://streamlit.io/)** existing app and added his **PANN predictive model** and some plots from his H2o wave app. \n\nYou can find the app live [here](https://share.streamlit.io/yassinealouini/rfcx-species-audio-detection-streamlit/app.py) and here is a screenshot: \n\n\n\n\n\nLet me know what you think and which features you would like to see.\n\n\nSome links:\n\n- The original app source code [repo](https://github.com/gaborfodor/wave-bird-recognition)\n- The streamlit app source code [repo](https://github.com/yassineAlouini/rfcx-species-audio-detection-streamlit)", "1337651": "It looks great! \nI tried it out with a sample 60-sec clip from Rainforest audio competition, but the app crashed by the way I hear the output audio (FYI the predictions has been returned):\n```\nOh no.\n\nError running app. If this keeps happening, please contact support.\n```", "1337674": "Thanks for giving it a try. Can you share your file please so that I can try to debug it. 👌", "1337678": "yassinealouini \nI picked a .flac file from here: \nhttps://www.kaggle.com/alexlwh/rfcx-small-train-data?select=train_small", "1337679": "I will give it a try. It might be the file format. I will post my findings. 👌", "1337690": "Alright, I had to reboot the app since it went down. \n\nThen, I tried one `.flac` file and it worked. \n\n\n\n\nCould you retry @alexlwh please? By the way, can you share your browser version and type? Thanks for helping me investigate. :)", "1337699": "One other hypothesis I have is that due to how the application is deployed, it could be rebooting each time too much traffic is reached. 😄", "1337702": "So one solution is to get the app's source code and deploy it on your own. It is very easy using the streamlit deploy documentation [page](https://docs.streamlit.io/en/stable/deploy_streamlit_app.html).", "1337981": "Perfect! it works for me now\nIt looks pretty nice 👍", "1337982": "What did u use for deployment btw?", "1338044": "I am using the streamlit's own infra but you can do it with many different providers. As said above, the \ndeploy [page](https://docs.streamlit.io/en/stable/deploy_streamlit_app.html) is very helpful. \n\nBasically, you need:\n\n- sign up to streamlit (using Github is the easisest way)\n- create a Github repo (should be public to have access to the free tier)\n- deploy, you have 3 free apps\n\nIf you have trouble doing it, let me know." }, "source": "meta" }