{ "id": 581187, "title": "Stop to share any high score public notebook in the last week", "url": "/competitions/birdclef-2025/discussion/581187", "author_name": "", "post_date": "2025-05-28T23:17:53.771689600Z", "votes": 31, "comment_count": 10, "views": 0, "content": "

Hey, just a heads up — as per usual Kaggle etiquette, let's avoid posting top-scoring code after 23:59. We've always done that in past comps, right?

\n

May 29, 2025 - Entry Deadline. You must accept the competition rules before this date to compete.

\n

May 29, 2025 - Team Merger Deadline. This is the last day participants may join or merge teams.

", "messages": [ { "id": "3211830", "postDate": "05/28/2025 23:17:53", "content": "

Hey, just a heads up — as per usual Kaggle etiquette, let's avoid posting top-scoring code after 23:59. We've always done that in past comps, right?

\n

May 29, 2025 - Entry Deadline. You must accept the competition rules before this date to compete.

\n

May 29, 2025 - Team Merger Deadline. This is the last day participants may join or merge teams.

", "rawMarkdown": "Hey, just a heads up — as per usual Kaggle etiquette, let's avoid posting top-scoring code after 23:59. We've always done that in past comps, right?\n\nMay 29, 2025 - Entry Deadline. You must accept the competition rules before this date to compete.\n\nMay 29, 2025 - Team Merger Deadline. This is the last day participants may join or merge teams.", "votes": null }, { "id": "3211920", "postDate": "05/29/2025 04:00:41", "content": "

Even you know, what's going to happen, 4 days before deadline 🥲.
\nKindly please don't share it, you can get notebook medal later too.

", "rawMarkdown": "Even you know, what's going to happen, 4 days before deadline 🥲. \nKindly please don't share it, you can get notebook medal later too.", "votes": null }, { "id": "3211921", "postDate": "05/29/2025 04:01:23", "content": "

I think Kaggle rules anyway don't allow sharing high scoring public content after the merger deadline @siyingya0

", "rawMarkdown": "I think Kaggle rules anyway don't allow sharing high scoring public content after the merger deadline @siyingya0", "votes": null }, { "id": "3211938", "postDate": "05/29/2025 04:41:19", "content": "

I think the truly qualified players in Kaggle don't rely on submitting public notebook to get a medal.

", "rawMarkdown": "I think the truly qualified players in Kaggle don't rely on submitting public notebook to get a medal.", "votes": null }, { "id": "3211951", "postDate": "05/29/2025 05:13:40", "content": "

Yes, as a Kaggle participant, we have always followed the rules — if not for some participants who choose to break them.

", "rawMarkdown": "Yes, as a Kaggle participant, we have always followed the rules — if not for some participants who choose to break them.", "votes": null }, { "id": "3211953", "postDate": "05/29/2025 05:16:45", "content": "

This is an unwritten rule that all participants are expected to follow. Please don't try to argue against it or break it.

", "rawMarkdown": "This is an unwritten rule that all participants are expected to follow. Please don't try to argue against it or break it.", "votes": null }, { "id": "3211954", "postDate": "05/29/2025 05:18:29", "content": "

The purpose of this post is to remind some newer Kaggle participants.

", "rawMarkdown": "The purpose of this post is to remind some newer Kaggle participants.", "votes": null }, { "id": "3212306", "postDate": "05/29/2025 15:56:00", "content": "

i think kaggle don'tnallow this

", "rawMarkdown": "i think kaggle don'tnallow this", "votes": null }, { "id": "3213906", "postDate": "05/30/2025 16:32:59", "content": "

From the perspective of a newbie like me (definitely not gonna get qualified). I managed to get in top 100 last week from scratch, and losing hundreds of places because hundreds of people just copied-pasted is quite discouraging

", "rawMarkdown": "From the perspective of a newbie like me (definitely not gonna get qualified). I managed to get in top 100 last week from scratch, and losing hundreds of places because hundreds of people just copied-pasted is quite discouraging", "votes": null }, { "id": "3214027", "postDate": "05/30/2025 19:32:11", "content": "

Many high-scoring public notebooks are likely overfit to the public data with specific tweaks, like the function that applies a power to less prominent columns, so those who just copy them might not hold their rank on the private leaderboard.
\nMy own model, trained from scratch, reached 0.878 days ago, and I'm hoping that by not over-adjusting it to the public set, it will perform well against those recently shared.
\nKeep working on your own model; it can definitely shine in the end!

", "rawMarkdown": "Many high-scoring public notebooks are likely overfit to the public data with specific tweaks, like the function that applies a power to less prominent columns, so those who just copy them might not hold their rank on the private leaderboard.\nMy own model, trained from scratch, reached 0.878 days ago, and I'm hoping that by not over-adjusting it to the public set, it will perform well against those recently shared.\nKeep working on your own model; it can definitely shine in the end!", "votes": null }, { "id": "3215373", "postDate": "06/02/2025 05:57:26", "content": "

Well going by the past Birdclef competition trends people just blindly forking and sharing using public kernels are prone to shake down on the leaderboard when results of private test is reveal so overall the models overfitting public test would get punished .
\nNonetheless there should be some steps or measures taken by Kaggle to avoid users sharing high scoring notebook after merger deadline is out

", "rawMarkdown": "Well going by the past Birdclef competition trends people just blindly forking and sharing using public kernels are prone to shake down on the leaderboard when results of private test is reveal so overall the models overfitting public test would get punished .\nNonetheless there should be some steps or measures taken by Kaggle to avoid users sharing high scoring notebook after merger deadline is out", "votes": null } ], "comments": [ { "id": 3211920, "author_name": "arpit1bansal", "author_url": "", "post_date": "05/29/2025 04:00:41", "content": "

Even you know, what's going to happen, 4 days before deadline 🥲.
\nKindly please don't share it, you can get notebook medal later too.

", "votes": null, "replies": [] }, { "id": 3211921, "author_name": "ravi20076", "author_url": "", "post_date": "05/29/2025 04:01:23", "content": "

I think Kaggle rules anyway don't allow sharing high scoring public content after the merger deadline @siyingya0

", "votes": null, "replies": [ { "id": 3211951, "author_name": "siyingya0", "author_url": "", "post_date": "05/29/2025 05:13:40", "content": "

Yes, as a Kaggle participant, we have always followed the rules — if not for some participants who choose to break them.

", "votes": null, "replies": [] } ] }, { "id": 3211938, "author_name": "yutongzhang20080108", "author_url": "", "post_date": "05/29/2025 04:41:19", "content": "

I think the truly qualified players in Kaggle don't rely on submitting public notebook to get a medal.

", "votes": null, "replies": [ { "id": 3211953, "author_name": "siyingya0", "author_url": "", "post_date": "05/29/2025 05:16:45", "content": "

This is an unwritten rule that all participants are expected to follow. Please don't try to argue against it or break it.

", "votes": null, "replies": [] }, { "id": 3213906, "author_name": "antoinemasq", "author_url": "", "post_date": "05/30/2025 16:32:59", "content": "

From the perspective of a newbie like me (definitely not gonna get qualified). I managed to get in top 100 last week from scratch, and losing hundreds of places because hundreds of people just copied-pasted is quite discouraging

", "votes": null, "replies": [ { "id": 3214027, "author_name": "carloscalvinp", "author_url": "", "post_date": "05/30/2025 19:32:11", "content": "

Many high-scoring public notebooks are likely overfit to the public data with specific tweaks, like the function that applies a power to less prominent columns, so those who just copy them might not hold their rank on the private leaderboard.
\nMy own model, trained from scratch, reached 0.878 days ago, and I'm hoping that by not over-adjusting it to the public set, it will perform well against those recently shared.
\nKeep working on your own model; it can definitely shine in the end!

", "votes": null, "replies": [] } ] } ] }, { "id": 3211954, "author_name": "siyingya0", "author_url": "", "post_date": "05/29/2025 05:18:29", "content": "

The purpose of this post is to remind some newer Kaggle participants.

", "votes": null, "replies": [] }, { "id": 3212306, "author_name": "khankhantmg", "author_url": "", "post_date": "05/29/2025 15:56:00", "content": "

i think kaggle don'tnallow this

", "votes": null, "replies": [] }, { "id": 3215373, "author_name": "sayedathar11", "author_url": "", "post_date": "06/02/2025 05:57:26", "content": "

Well going by the past Birdclef competition trends people just blindly forking and sharing using public kernels are prone to shake down on the leaderboard when results of private test is reveal so overall the models overfitting public test would get punished .
\nNonetheless there should be some steps or measures taken by Kaggle to avoid users sharing high scoring notebook after merger deadline is out

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "3211830": "Hey, just a heads up — as per usual Kaggle etiquette, let's avoid posting top-scoring code after 23:59. We've always done that in past comps, right?\n\nMay 29, 2025 - Entry Deadline. You must accept the competition rules before this date to compete.\n\nMay 29, 2025 - Team Merger Deadline. This is the last day participants may join or merge teams.", "3211920": "Even you know, what's going to happen, 4 days before deadline 🥲. \nKindly please don't share it, you can get notebook medal later too.", "3211921": "I think Kaggle rules anyway don't allow sharing high scoring public content after the merger deadline @siyingya0", "3211938": "I think the truly qualified players in Kaggle don't rely on submitting public notebook to get a medal.", "3211951": "Yes, as a Kaggle participant, we have always followed the rules — if not for some participants who choose to break them.", "3211953": "This is an unwritten rule that all participants are expected to follow. Please don't try to argue against it or break it.", "3211954": "The purpose of this post is to remind some newer Kaggle participants.", "3212306": "i think kaggle don'tnallow this", "3213906": "From the perspective of a newbie like me (definitely not gonna get qualified). I managed to get in top 100 last week from scratch, and losing hundreds of places because hundreds of people just copied-pasted is quite discouraging", "3214027": "Many high-scoring public notebooks are likely overfit to the public data with specific tweaks, like the function that applies a power to less prominent columns, so those who just copy them might not hold their rank on the private leaderboard.\nMy own model, trained from scratch, reached 0.878 days ago, and I'm hoping that by not over-adjusting it to the public set, it will perform well against those recently shared.\nKeep working on your own model; it can definitely shine in the end!", "3215373": "Well going by the past Birdclef competition trends people just blindly forking and sharing using public kernels are prone to shake down on the leaderboard when results of private test is reveal so overall the models overfitting public test would get punished .\nNonetheless there should be some steps or measures taken by Kaggle to avoid users sharing high scoring notebook after merger deadline is out" }, "source": "meta" }