Submission CSV not found error. The kernel ran fine though and I can see the submission.csv file in the output tab. Anyone else facing the same issue?\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F536977%2F813ffcb37c82c084c54bc2d0b907547f%2FScreen%20Shot%202019-08-13%20at%2012.24.50%20AM.png?generation=1565636210980751&alt=media\")
I tried to submit to the comp twice but got Submission CSV not found error. The kernel ran fine though and I can see the submission.csv file in the output tab. Anyone else facing the same issue?\n
I tried to submit to the comp twice but got Submission CSV not found error. The kernel ran fine though and I can see the submission.csv file in the output tab. Anyone else facing the same issue?\n
Submission CSV Not Found indicates a problem with your code; everything ran fine but our scoring system wasn't able to find what it expected. Since the private set is different from the public set, being able to see the submission.csv on your output tab does not mean your kernel works when it gets run again for a score.
I don't see any problem in my code. May be it is better to provide the logs for the error so that people know what went wrong. If the kernel ran fine and if there is something wrong with the test run, how do you expect people to debug it without logs?
In the past few weeks, people were saying that Submission CSV Not Found error is most likely caused by memory overflow. Hope this can help you.
I suspected that and reduced my batch size. In fact, I even tried with batch size of 1 but no success
Thank you for sharing! Also been having issues in the past few days. Trying to reduce the \"pressure\" on RAM now, but I'm not sure what is wrong. Is it likely due to memory overflow?
\n\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F1956324%2F73529f02ff042ffa2911b18ef16d9c77%2Fissue.png?generation=1565705363253321&alt=media\")
Excellent idea! The overall experience of synchronous kernel competition has been annoying so far and providing logs would greatly save us time debugging for the test run.
", "rawMarkdown": "Excellent idea! The overall experience of synchronous kernel competition has been annoying so far and providing logs would greatly save us time debugging for the test run.", "votes": null }, { "id": "598444", "postDate": "08/13/2019 15:27:59", "content": "Suspecting of memory overflow issues, I reduced my batch size to the least but it didn't help. The problem is that without error logs there is no way to debug it. I am seriously missing the point of not providing error logs if there is some error
", "rawMarkdown": "Suspecting of memory overflow issues, I reduced my batch size to the least but it didn't help. The problem is that without error logs there is no way to debug it. I am seriously missing the point of not providing error logs if there is some error", "votes": null }, { "id": "598500", "postDate": "08/13/2019 16:43:31", "content": "If we provided raw error logs, we wouldn't be able to maintain the security of the private test set.
", "rawMarkdown": "If we provided raw error logs, we wouldn't be able to maintain the security of the private test set.", "votes": null }, { "id": "598766", "postDate": "08/14/2019 02:27:32", "content": "Have you take a look at this 5 common sources yet?
\n\nhttps://www.kaggle.com/c/aptos2019-blindness-detection/discussion/99227#latest-577724
", "rawMarkdown": "Have you take a look at this 5 common sources yet?\n\nhttps://www.kaggle.com/c/aptos2019-blindness-detection/discussion/99227#latest-577724", "votes": null }, { "id": "615289", "postDate": "09/01/2019 18:23:53", "content": "HI Sohier Dane ,\nI am also facing same issue ,i have submitted 5 times and getting same error.\nCan you please suggest what to do now?\nWe should definitely know actual issues as deadline of submission is very close and today after 5 tries i couldnt submit also.
", "rawMarkdown": "HI Sohier Dane ,\nI am also facing same issue ,i have submitted 5 times and getting same error.\nCan you please suggest what to do now?\nWe should definitely know actual issues as deadline of submission is very close and today after 5 tries i couldnt submit also.", "votes": null } ], "comments": [ { "id": 597856, "author_name": "sohier", "author_url": "", "post_date": "08/12/2019 20:12:32", "content": "Submission CSV Not Found indicates a problem with your code; everything ran fine but our scoring system wasn't able to find what it expected. Since the private set is different from the public set, being able to see the submission.csv on your output tab does not mean your kernel works when it gets run again for a score.
I don't see any problem in my code. May be it is better to provide the logs for the error so that people know what went wrong. If the kernel ran fine and if there is something wrong with the test run, how do you expect people to debug it without logs?
Excellent idea! The overall experience of synchronous kernel competition has been annoying so far and providing logs would greatly save us time debugging for the test run.
", "votes": null, "replies": [] }, { "id": 615289, "author_name": "rajnijain7782", "author_url": "", "post_date": "09/01/2019 18:23:53", "content": "HI Sohier Dane ,\nI am also facing same issue ,i have submitted 5 times and getting same error.\nCan you please suggest what to do now?\nWe should definitely know actual issues as deadline of submission is very close and today after 5 tries i couldnt submit also.
", "votes": null, "replies": [] } ] }, { "id": 598200, "author_name": "haqishen", "author_url": "", "post_date": "08/13/2019 09:05:47", "content": "In the past few weeks, people were saying that Submission CSV Not Found error is most likely caused by memory overflow. Hope this can help you.
I suspected that and reduced my batch size. In fact, I even tried with batch size of 1 but no success
Thank you for sharing! Also been having issues in the past few days. Trying to reduce the \"pressure\" on RAM now, but I'm not sure what is wrong. Is it likely due to memory overflow?
\n\n
Suspecting of memory overflow issues, I reduced my batch size to the least but it didn't help. The problem is that without error logs there is no way to debug it. I am seriously missing the point of not providing error logs if there is some error
", "votes": null, "replies": [] }, { "id": 598500, "author_name": "sohier", "author_url": "", "post_date": "08/13/2019 16:43:31", "content": "If we provided raw error logs, we wouldn't be able to maintain the security of the private test set.
", "votes": null, "replies": [] } ] }, { "id": 598766, "author_name": "ratthachat", "author_url": "", "post_date": "08/14/2019 02:27:32", "content": "Have you take a look at this 5 common sources yet?
\n\nhttps://www.kaggle.com/c/aptos2019-blindness-detection/discussion/99227#latest-577724
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "597812": "I tried to submit to the comp twice but got `Submission CSV not found` error. The kernel ran fine though and I can see the `submission.csv` file in the `output` tab. Anyone else facing the same issue?\n", "597856": "`Submission CSV Not Found` indicates a problem with your code; everything ran fine but our scoring system wasn't able to find what it expected. Since the private set is different from the public set, being able to see the `submission.csv` on your output tab does not mean your kernel works when it gets run again for a score.", "598104": "I don't see any problem in my code. May be it is better to provide the logs for the error so that people know what went wrong. If the kernel ran fine and if there is something wrong with the `test run`, how do you expect people to debug it without logs?", "598200": "In the past few weeks, people were saying that `Submission CSV Not Found` error is most likely caused by memory overflow. Hope this can help you.", "598220": "I suspected that and reduced my batch size. In fact, I even tried with batch size of `1` but no success", "598364": "Thank you for sharing! Also been having issues in the past few days. Trying to reduce the \"pressure\" on RAM now, but I'm not sure what is wrong. Is it likely due to memory overflow?\n\n", "598367": "Excellent idea! The overall experience of synchronous kernel competition has been annoying so far and providing logs would greatly save us time debugging for the test run.", "598444": "Suspecting of memory overflow issues, I reduced my batch size to the least but it didn't help. The problem is that without error logs there is no way to debug it. I am seriously missing the point of not providing error logs if there is some error", "598500": "If we provided raw error logs, we wouldn't be able to maintain the security of the private test set.", "598766": "Have you take a look at this 5 common sources yet?\n\nhttps://www.kaggle.com/c/aptos2019-blindness-detection/discussion/99227#latest-577724", "615289": "HI Sohier Dane ,\nI am also facing same issue ,i have submitted 5 times and getting same error.\nCan you please suggest what to do now?\nWe should definitely know actual issues as deadline of submission is very close and today after 5 tries i couldnt submit also." }, "source": "meta" }