Hi, I have managed to download the data for this competition through the browser with no problem. When I try to do it via the commandline, however, it doesn't seem to work. The process just fills my ram until my vps freezes but writes nothing to disk. I have also tried to make a 240gb swap file. I also fills this and then nothing seems to be happening. No output from the cli-tool. As a sanity check, I have also tried it with another competition dataset (code golf) and there it works fine. Has anyone else encountered this?
", "messages": [ { "id": 3298420, "postDate": "2025-10-05T13:55:38.743Z", "content": "Another kaggle version may resolve this issue.
\n(pip install kaggle==1.5.16
\nhttps://www.kaggle.com/competitions/rsna-intracranial-aneurysm-detection/discussion/593731
Yea you need to use that specific version of kaggle for some reaeson
", "rawMarkdown": "Yea you need to use that specific version of kaggle for some reaeson", "votes": 1, "replies": [ { "id": 3298436, "postDate": "2025-10-05T14:41:17.233Z", "content": "Thanks! Now it works. That was an annoying way to send several hours :)
", "rawMarkdown": "Thanks! Now it works. That was an annoying way to send several hours :)" } ] } ] }, { "id": 3298419, "postDate": "2025-10-05T13:54:28.077Z", "content": "Yep had same issues, even with some other datasets. Using kagglehub worked for me instead of using the default kaggle api. (It also resumes downloads if execution stops midway)
", "rawMarkdown": "Yep had same issues, even with some other datasets. Using kagglehub worked for me instead of using the default kaggle api. (It also resumes downloads if execution stops midway)", "votes": 1 }, { "id": 3298503, "postDate": "2025-10-05T17:23:15.037Z", "content": "Hi Thomas, I haven’t faced this issue myself, but it sounds like the CLI might be struggling with the large dataset size. A few things you could try:
\nMake sure your CLI tool is updated to the latest version.
\nTry downloading in smaller chunks if the CLI supports it.
\nCheck if there’s enough disk space allocated and not just RAM/swap.
\nAs a workaround, you could continue using the browser download or a system with more RAM for the initial download.
\nHas anyone else experienced this issue with the RSNA dataset via CLI?
", "rawMarkdown": "Hi Thomas, I haven’t faced this issue myself, but it sounds like the CLI might be struggling with the large dataset size. A few things you could try:\n\nMake sure your CLI tool is updated to the latest version.\n\nTry downloading in smaller chunks if the CLI supports it.\n\nCheck if there’s enough disk space allocated and not just RAM/swap.\n\nAs a workaround, you could continue using the browser download or a system with more RAM for the initial download.\n\nHas anyone else experienced this issue with the RSNA dataset via CLI?", "votes": 2, "replies": [ { "id": 3298542, "postDate": "2025-10-05T19:22:32.380Z", "content": "Thanks for you reply! I had basically tried everything that I (and ChatGPT) could come up with. In the end it was really just the version of the kaggle cli that seems to have some issues. pip install kaggle==1.5.16 fixed it!
", "rawMarkdown": "Thanks for you reply! I had basically tried everything that I (and ChatGPT) could come up with. In the end it was really just the version of the kaggle cli that seems to have some issues. pip install kaggle==1.5.16 fixed it!" } ] }, { "id": 3298409, "postDate": "2025-10-05T13:27:17.513Z", "content": "Hi, I have managed to download the data for this competition through the browser with no problem. When I try to do it via the commandline, however, it doesn't seem to work. The process just fills my ram until my vps freezes but writes nothing to disk. I have also tried to make a 240gb swap file. I also fills this and then nothing seems to be happening. No output from the cli-tool. As a sanity check, I have also tried it with another competition dataset (code golf) and there it works fine. Has anyone else encountered this?
", "rawMarkdown": "Hi, I have managed to download the data for this competition through the browser with no problem. When I try to do it via the commandline, however, it doesn't seem to work. The process just fills my ram until my vps freezes but writes nothing to disk. I have also tried to make a 240gb swap file. I also fills this and then nothing seems to be happening. No output from the cli-tool. As a sanity check, I have also tried it with another competition dataset (code golf) and there it works fine. Has anyone else encountered this? " } ], "comments": [ { "id": 3298420, "author_name": "shiba-inu", "author_url": "", "post_date": "2025-10-05T13:55:38.743000", "content": "Another kaggle version may resolve this issue.
\n(pip install kaggle==1.5.16
\nhttps://www.kaggle.com/competitions/rsna-intracranial-aneurysm-detection/discussion/593731
Yea you need to use that specific version of kaggle for some reaeson
", "votes": 1, "replies": [ { "id": 3298436, "author_name": "thomas rost", "author_url": "", "post_date": "2025-10-05T14:41:17.233000", "content": "Thanks! Now it works. That was an annoying way to send several hours :)
", "votes": 0, "replies": [] } ] } ] }, { "id": 3298419, "author_name": "Satwik", "author_url": "", "post_date": "2025-10-05T13:54:28.077000", "content": "Yep had same issues, even with some other datasets. Using kagglehub worked for me instead of using the default kaggle api. (It also resumes downloads if execution stops midway)
", "votes": 1, "replies": [] }, { "id": 3298503, "author_name": "Prince Rajak", "author_url": "", "post_date": "2025-10-05T17:23:15.037000", "content": "Hi Thomas, I haven’t faced this issue myself, but it sounds like the CLI might be struggling with the large dataset size. A few things you could try:
\nMake sure your CLI tool is updated to the latest version.
\nTry downloading in smaller chunks if the CLI supports it.
\nCheck if there’s enough disk space allocated and not just RAM/swap.
\nAs a workaround, you could continue using the browser download or a system with more RAM for the initial download.
\nHas anyone else experienced this issue with the RSNA dataset via CLI?
", "votes": 2, "replies": [ { "id": 3298542, "author_name": "thomas rost", "author_url": "", "post_date": "2025-10-05T19:22:32.380000", "content": "Thanks for you reply! I had basically tried everything that I (and ChatGPT) could come up with. In the end it was really just the version of the kaggle cli that seems to have some issues. pip install kaggle==1.5.16 fixed it!
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "3298420": "Another kaggle version may resolve this issue.\n(pip install kaggle==1.5.16\nhttps://www.kaggle.com/competitions/rsna-intracranial-aneurysm-detection/discussion/593731\n", "3298419": "Yep had same issues, even with some other datasets. Using kagglehub worked for me instead of using the default kaggle api. (It also resumes downloads if execution stops midway)", "3298503": "Hi Thomas, I haven’t faced this issue myself, but it sounds like the CLI might be struggling with the large dataset size. A few things you could try:\n\nMake sure your CLI tool is updated to the latest version.\n\nTry downloading in smaller chunks if the CLI supports it.\n\nCheck if there’s enough disk space allocated and not just RAM/swap.\n\nAs a workaround, you could continue using the browser download or a system with more RAM for the initial download.\n\nHas anyone else experienced this issue with the RSNA dataset via CLI?", "3298409": "Hi, I have managed to download the data for this competition through the browser with no problem. When I try to do it via the commandline, however, it doesn't seem to work. The process just fills my ram until my vps freezes but writes nothing to disk. I have also tried to make a 240gb swap file. I also fills this and then nothing seems to be happening. No output from the cli-tool. As a sanity check, I have also tried it with another competition dataset (code golf) and there it works fine. Has anyone else encountered this? " } }