I've been trying to download the competition data locally, via kaggle competitions download -c quickdraw-doodle-recognition, but I'm not having a lot of luck - train_raw.zip is coming down at somewhere around 300 kb/sec, and at 65.9 G that's going to take days. My local connection is nowhere near this slow: Google's speed test clocks me at 97 megabits/sec, and even in this download set, all the CSV files come down at around 10 MB/sec, 30x faster than the big file.
Is there another way I should be downloading this stuff, hitting a different endpoint maybe? I'm brand new to Kaggle, so it's likely I'm doing something wrong.
", "messages": [ { "id": 399107, "postDate": "2018-10-05T08:07:07.910Z", "content": "I've been trying to download the competition data locally, via kaggle competitions download -c quickdraw-doodle-recognition, but I'm not having a lot of luck - train_raw.zip is coming down at somewhere around 300 kb/sec, and at 65.9 G that's going to take days. My local connection is nowhere near this slow: Google's speed test clocks me at 97 megabits/sec, and even in this download set, all the CSV files come down at around 10 MB/sec, 30x faster than the big file.
Is there another way I should be downloading this stuff, hitting a different endpoint maybe? I'm brand new to Kaggle, so it's likely I'm doing something wrong.
", "rawMarkdown": "I've been trying to download the competition data locally, via `kaggle competitions download -c quickdraw-doodle-recognition`, but I'm not having a lot of luck - `train_raw.zip` is coming down at somewhere around 300 kb/sec, and at 65.9 G that's going to take days. My local connection is nowhere near this slow: Google's speed test clocks me at 97 megabits/sec, and even in this download set, all the CSV files come down at around 10 MB/sec, 30x faster than the big file.\n\nIs there another way I should be downloading this stuff, hitting a different endpoint maybe? I'm brand new to Kaggle, so it's likely I'm doing something wrong.", "votes": -1 }, { "id": 399110, "postDate": "2018-10-05T08:16:48.380Z", "content": "Ok, it looks like the web download runs fine at full speed (from the data tab), so I'm unblocked, but I'm still curious to understand why the command line download is so throttled, if anyone knows.
", "rawMarkdown": "Ok, it looks like the web download runs fine at full speed (from the data tab), so I'm unblocked, but I'm still curious to understand why the command line download is so throttled, if anyone knows." } ], "comments": [ { "id": 399110, "author_name": "Eric Jordan", "author_url": "", "post_date": "2018-10-05T08:16:48.380000", "content": "Ok, it looks like the web download runs fine at full speed (from the data tab), so I'm unblocked, but I'm still curious to understand why the command line download is so throttled, if anyone knows.
", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "399107": "I've been trying to download the competition data locally, via `kaggle competitions download -c quickdraw-doodle-recognition`, but I'm not having a lot of luck - `train_raw.zip` is coming down at somewhere around 300 kb/sec, and at 65.9 G that's going to take days. My local connection is nowhere near this slow: Google's speed test clocks me at 97 megabits/sec, and even in this download set, all the CSV files come down at around 10 MB/sec, 30x faster than the big file.\n\nIs there another way I should be downloading this stuff, hitting a different endpoint maybe? I'm brand new to Kaggle, so it's likely I'm doing something wrong.", "399110": "Ok, it looks like the web download runs fine at full speed (from the data tab), so I'm unblocked, but I'm still curious to understand why the command line download is so throttled, if anyone knows." } }