Guys, is it alright?\nI am quite new in Kaggle, never meet the condition in previous submissions. \nupdate: one submission goes timeout error.\nupdate2: I mean prediction CSV files submitted, and in status 'running' for an extremely long time.\nFixed by turned on GPU while submitting or simply just fixed by some unknown reasons.
", "messages": [ { "id": "763791", "postDate": "03/04/2020 21:29:40", "content": "Guys, is it alright?\nI am quite new in Kaggle, never meet the condition in previous submissions. \nupdate: one submission goes timeout error.\nupdate2: I mean prediction CSV files submitted, and in status 'running' for an extremely long time.\nFixed by turned on GPU while submitting or simply just fixed by some unknown reasons.
", "rawMarkdown": "Guys, is it alright?\nI am quite new in Kaggle, never meet the condition in previous submissions. \nupdate: one submission goes timeout error.\nupdate2: I mean prediction CSV files submitted, and in status 'running' for an extremely long time.\nFixed by turned on GPU while submitting or simply just fixed by some unknown reasons.", "votes": null }, { "id": "763854", "postDate": "03/04/2020 22:49:00", "content": "I saw somewhere in the discussion that the submission should be done in 2 hours. If your submission takes so long, it wouldn't be a valid one. I suggest you train the model in a different kernel. The submission kernel does prediction only, a typical model used in this competition takes 8-10 mins to run. More model you used, more time the submission takes.
", "rawMarkdown": "I saw somewhere in the discussion that the submission should be done in 2 hours. If your submission takes so long, it wouldn't be a valid one. I suggest you train the model in a different kernel. The submission kernel does prediction only, a typical model used in this competition takes 8-10 mins to run. More model you used, more time the submission takes.", "votes": null }, { "id": "763878", "postDate": "03/04/2020 23:33:08", "content": "Thanks for your reply, I did submission with the kernel that does prediction only. \nActually the CSV files were done and submitted for calculation which should be quite fast for this competition.
", "rawMarkdown": "Thanks for your reply, I did submission with the kernel that does prediction only. \nActually the CSV files were done and submitted for calculation which should be quite fast for this competition.", "votes": null }, { "id": "763881", "postDate": "03/04/2020 23:46:33", "content": "I see, then the submission won't be too long. In my case, one model takes about 8 mins. Btw, even you've generated the CSV file, when you submit the file, the calculation will be done again with its public test set. Thus, it takes a little time. :P For me, the whole process takes 20-30 mins, since I ensemble several models.
", "rawMarkdown": "I see, then the submission won't be too long. In my case, one model takes about 8 mins. Btw, even you've generated the CSV file, when you submit the file, the calculation will be done again with its public test set. Thus, it takes a little time. :P For me, the whole process takes 20-30 mins, since I ensemble several models.", "votes": null }, { "id": "765448", "postDate": "03/06/2020 16:25:17", "content": "I think there must be quite a load of computation being done at the moment, since the accelerator quotas are about to reset. So maybe becaude everyone is computing, submissions are getting delayed/queued?
", "rawMarkdown": "I think there must be quite a load of computation being done at the moment, since the accelerator quotas are about to reset. So maybe becaude everyone is computing, submissions are getting delayed/queued?", "votes": null } ], "comments": [ { "id": 763854, "author_name": "yuanlin08", "author_url": "", "post_date": "03/04/2020 22:49:00", "content": "I saw somewhere in the discussion that the submission should be done in 2 hours. If your submission takes so long, it wouldn't be a valid one. I suggest you train the model in a different kernel. The submission kernel does prediction only, a typical model used in this competition takes 8-10 mins to run. More model you used, more time the submission takes.
", "votes": null, "replies": [ { "id": 763878, "author_name": "zlanan", "author_url": "", "post_date": "03/04/2020 23:33:08", "content": "Thanks for your reply, I did submission with the kernel that does prediction only. \nActually the CSV files were done and submitted for calculation which should be quite fast for this competition.
", "votes": null, "replies": [] }, { "id": 763881, "author_name": "yuanlin08", "author_url": "", "post_date": "03/04/2020 23:46:33", "content": "I see, then the submission won't be too long. In my case, one model takes about 8 mins. Btw, even you've generated the CSV file, when you submit the file, the calculation will be done again with its public test set. Thus, it takes a little time. :P For me, the whole process takes 20-30 mins, since I ensemble several models.
", "votes": null, "replies": [] }, { "id": 765448, "author_name": "carlosmir", "author_url": "", "post_date": "03/06/2020 16:25:17", "content": "I think there must be quite a load of computation being done at the moment, since the accelerator quotas are about to reset. So maybe becaude everyone is computing, submissions are getting delayed/queued?
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "763791": "Guys, is it alright?\nI am quite new in Kaggle, never meet the condition in previous submissions. \nupdate: one submission goes timeout error.\nupdate2: I mean prediction CSV files submitted, and in status 'running' for an extremely long time.\nFixed by turned on GPU while submitting or simply just fixed by some unknown reasons.", "763854": "I saw somewhere in the discussion that the submission should be done in 2 hours. If your submission takes so long, it wouldn't be a valid one. I suggest you train the model in a different kernel. The submission kernel does prediction only, a typical model used in this competition takes 8-10 mins to run. More model you used, more time the submission takes.", "763878": "Thanks for your reply, I did submission with the kernel that does prediction only. \nActually the CSV files were done and submitted for calculation which should be quite fast for this competition.", "763881": "I see, then the submission won't be too long. In my case, one model takes about 8 mins. Btw, even you've generated the CSV file, when you submit the file, the calculation will be done again with its public test set. Thus, it takes a little time. :P For me, the whole process takes 20-30 mins, since I ensemble several models.", "765448": "I think there must be quite a load of computation being done at the moment, since the accelerator quotas are about to reset. So maybe becaude everyone is computing, submissions are getting delayed/queued?" }, "source": "meta" }