![\"Check](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F9733051%2F7c3a8b99727f1c27ee94102a39e5194d%2FCapture.JPG?generation=1675581216197951&alt=media\")
I fine tuned a pretrained resnet model, it ran sucessfully on test data. But the submission got failed, telling notebook threw exception. Can anyone please help me on this.
", "messages": [ { "id": 2131706, "postDate": "2023-02-06T09:56:48.450Z", "content": "Hello
\nI have the same problem. My code runs fine, generates a submission.csv in the precise format required, but somehow fails to submit. I've tried all the suggestions I've found on this forum (e.g., making sure the working folder is totally empty before saving the submission.csv). I'm using the pydicom code shared on this forum to process the source images. This works fine on all the training images and on the small (4 image) test set. I therefore assume it works on the true test set, but of course I cannot tell since there is no diagnostic feedback provided (!). To eliminate potential sources of error, I even tried ignoring (commenting-out) my machine-learning results and instead prepared a column of random numbers (between 0 and 1) of the same height as the test set (whatever that may be, unknown to me -- I simply parse the true test set on the fly). Again, the code runs, but fails to submit. So, I'm at a bit of a loss since I'm unable to get to the interesting stage of tweaking my model for improvements. Instead I'm stuck trying to sort a frustrating data-wrangling issue…
\nMany thanks in advance
", "rawMarkdown": "Hello\n\nI have the same problem. My code runs fine, generates a submission.csv in the precise format required, but somehow fails to submit. I've tried all the suggestions I've found on this forum (e.g., making sure the working folder is totally empty before saving the submission.csv). I'm using the pydicom code shared on this forum to process the source images. This works fine on all the training images and on the small (4 image) test set. I therefore assume it works on the true test set, but of course I cannot tell since there is no diagnostic feedback provided (!). To eliminate potential sources of error, I even tried ignoring (commenting-out) my machine-learning results and instead prepared a column of random numbers (between 0 and 1) of the same height as the test set (whatever that may be, unknown to me -- I simply parse the true test set on the fly). Again, the code runs, but fails to submit. So, I'm at a bit of a loss since I'm unable to get to the interesting stage of tweaking my model for improvements. Instead I'm stuck trying to sort a frustrating data-wrangling issue...\n\nMany thanks in advance ", "votes": 1, "replies": [ { "id": 2131817, "postDate": "2023-02-06T12:16:04.420Z", "content": "I dont what is the matter with submission, because when I tried submitting 23 days ago after training a simple model it got submitted. But using the same notebook when I change epoch from 1 to 5 and then submitted it got failed.
", "rawMarkdown": "I dont what is the matter with submission, because when I tried submitting 23 days ago after training a simple model it got submitted. But using the same notebook when I change epoch from 1 to 5 and then submitted it got failed. " }, { "id": 2154945, "postDate": "2023-02-22T10:44:07.360Z", "content": "I completely understand. Is Kaggle the home of Data Science or just frustrating data wrangling?? It would be nice if the competitions, even though they are trying to be as realistic as possible, were a bit more setup for success???
", "rawMarkdown": "I completely understand. Is Kaggle the home of Data Science or just frustrating data wrangling?? It would be nice if the competitions, even though they are trying to be as realistic as possible, were a bit more setup for success???", "replies": [ { "id": 2154956, "postDate": "2023-02-22T10:50:58.707Z", "content": "Well said, Geoff. I have 19 deep learning models trained using TensorFlow for this competition. They run fine in my offline dev environment on the small 4-sample test set as well as on very large fictitious test sets I've crafted to try and resolve the submission issue. But I cannot get a submission to go through successfully. I can only guess that my image-processing is taking too long (via pydicom) -- but I simply do not know. Time spent on resolving this submission glitch is now quite a bit longer than time spend building the ML models. So, time to move on for now, I'm afraid…
", "rawMarkdown": "Well said, Geoff. I have 19 deep learning models trained using TensorFlow for this competition. They run fine in my offline dev environment on the small 4-sample test set as well as on very large fictitious test sets I've crafted to try and resolve the submission issue. But I cannot get a submission to go through successfully. I can only guess that my image-processing is taking too long (via pydicom) -- but I simply do not know. Time spent on resolving this submission glitch is now quite a bit longer than time spend building the ML models. So, time to move on for now, I'm afraid..." } ] } ] }, { "id": 2130107, "postDate": "2023-02-05T06:30:19.957Z", "content": "You need to submit as like format of sample submission.csv. Check sample submission.csv file.
", "rawMarkdown": "You need to submit as like format of sample submission.csv. Check sample submission.csv file.", "votes": 1, "replies": [ { "id": 2130125, "postDate": "2023-02-05T06:55:12.050Z", "content": "Thanks for replying,
\nI did the in the same format as sample submission.csv.
I fine tuned a pretrained resnet model, it ran sucessfully on test data. But the submission got failed, telling notebook threw exception. Can anyone please help me on this.
", "rawMarkdown": "I fine tuned a pretrained resnet model, it ran sucessfully on test data. But the submission got failed, telling notebook threw exception. Can anyone please help me on this. ", "votes": 2 }, { "id": 2151943, "postDate": "2023-02-20T13:33:15.620Z", "content": "is anyone able to solve this issue? I am unable to make even one submission.
", "rawMarkdown": "is anyone able to solve this issue? I am unable to make even one submission.", "replies": [ { "id": 2153814, "postDate": "2023-02-21T16:47:36.970Z", "content": "take reference
\nhttps://www.kaggle.com/code/markwijkhuizen/rsna-convnextv2-inference-tensorflow
Hello, have you figured out what is reason that caused submission problem?
", "rawMarkdown": "Hello, have you figured out what is reason that caused submission problem?\n", "replies": [ { "id": 2143447, "postDate": "2023-02-14T09:11:36.653Z", "content": "Unfortunately, not. I've had to move on to other things. Obviously others have figure out the submission process successfully, I could not : (
", "rawMarkdown": "Unfortunately, not. I've had to move on to other things. Obviously others have figure out the submission process successfully, I could not : (" }, { "id": 2143589, "postDate": "2023-02-14T11:20:32.863Z", "content": "Please look here, this may help.
", "rawMarkdown": "Please look here, this may help.", "replies": [ { "id": 2143590, "postDate": "2023-02-14T11:20:41.533Z", "content": "https://www.kaggle.com/code/markwijkhuizen/rsna-convnextv2-inference-tensorflow
", "rawMarkdown": "https://www.kaggle.com/code/markwijkhuizen/rsna-convnextv2-inference-tensorflow" }, { "id": 2154857, "postDate": "2023-02-22T09:34:18.003Z", "content": "Many thanks, very kind of your offer to help. I had a look at that example but the image processing would entail a rewrite of my (much simpler) pydicom approach so I think I'll have to sit this one out…
", "rawMarkdown": "Many thanks, very kind of your offer to help. I had a look at that example but the image processing would entail a rewrite of my (much simpler) pydicom approach so I think I'll have to sit this one out..." } ] } ] }, { "id": 2130136, "postDate": "2023-02-05T07:13:48.877Z", "content": "
Hello
\nI have the same problem. My code runs fine, generates a submission.csv in the precise format required, but somehow fails to submit. I've tried all the suggestions I've found on this forum (e.g., making sure the working folder is totally empty before saving the submission.csv). I'm using the pydicom code shared on this forum to process the source images. This works fine on all the training images and on the small (4 image) test set. I therefore assume it works on the true test set, but of course I cannot tell since there is no diagnostic feedback provided (!). To eliminate potential sources of error, I even tried ignoring (commenting-out) my machine-learning results and instead prepared a column of random numbers (between 0 and 1) of the same height as the test set (whatever that may be, unknown to me -- I simply parse the true test set on the fly). Again, the code runs, but fails to submit. So, I'm at a bit of a loss since I'm unable to get to the interesting stage of tweaking my model for improvements. Instead I'm stuck trying to sort a frustrating data-wrangling issue…
\nMany thanks in advance
", "votes": 1, "replies": [ { "id": 2131817, "author_name": "Shujaat Hasan", "author_url": "", "post_date": "2023-02-06T12:16:04.420000", "content": "I dont what is the matter with submission, because when I tried submitting 23 days ago after training a simple model it got submitted. But using the same notebook when I change epoch from 1 to 5 and then submitted it got failed.
", "votes": 0, "replies": [] }, { "id": 2154945, "author_name": "Geoff", "author_url": "", "post_date": "2023-02-22T10:44:07.360000", "content": "I completely understand. Is Kaggle the home of Data Science or just frustrating data wrangling?? It would be nice if the competitions, even though they are trying to be as realistic as possible, were a bit more setup for success???
", "votes": 0, "replies": [ { "id": 2154956, "author_name": "Yusuf Jafry", "author_url": "", "post_date": "2023-02-22T10:50:58.707000", "content": "Well said, Geoff. I have 19 deep learning models trained using TensorFlow for this competition. They run fine in my offline dev environment on the small 4-sample test set as well as on very large fictitious test sets I've crafted to try and resolve the submission issue. But I cannot get a submission to go through successfully. I can only guess that my image-processing is taking too long (via pydicom) -- but I simply do not know. Time spent on resolving this submission glitch is now quite a bit longer than time spend building the ML models. So, time to move on for now, I'm afraid…
", "votes": 0, "replies": [] } ] } ] }, { "id": 2130107, "author_name": "MD. RIPON MIAH", "author_url": "", "post_date": "2023-02-05T06:30:19.957000", "content": "You need to submit as like format of sample submission.csv. Check sample submission.csv file.
", "votes": 1, "replies": [ { "id": 2130125, "author_name": "Shujaat Hasan", "author_url": "", "post_date": "2023-02-05T06:55:12.050000", "content": "Thanks for replying,
\nI did the in the same format as sample submission.csv.
is anyone able to solve this issue? I am unable to make even one submission.
", "votes": 0, "replies": [ { "id": 2153814, "author_name": "Shujaat Hasan", "author_url": "", "post_date": "2023-02-21T16:47:36.970000", "content": "take reference
\nhttps://www.kaggle.com/code/markwijkhuizen/rsna-convnextv2-inference-tensorflow
Hello, have you figured out what is reason that caused submission problem?
", "votes": 0, "replies": [ { "id": 2143447, "author_name": "Yusuf Jafry", "author_url": "", "post_date": "2023-02-14T09:11:36.653000", "content": "Unfortunately, not. I've had to move on to other things. Obviously others have figure out the submission process successfully, I could not : (
", "votes": 0, "replies": [] }, { "id": 2143589, "author_name": "Shujaat Hasan", "author_url": "", "post_date": "2023-02-14T11:20:32.863000", "content": "Please look here, this may help.
", "votes": 0, "replies": [ { "id": 2143590, "author_name": "Shujaat Hasan", "author_url": "", "post_date": "2023-02-14T11:20:41.533000", "content": "https://www.kaggle.com/code/markwijkhuizen/rsna-convnextv2-inference-tensorflow
", "votes": 0, "replies": [] }, { "id": 2154857, "author_name": "Yusuf Jafry", "author_url": "", "post_date": "2023-02-22T09:34:18.003000", "content": "Many thanks, very kind of your offer to help. I had a look at that example but the image processing would entail a rewrite of my (much simpler) pydicom approach so I think I'll have to sit this one out…
", "votes": 0, "replies": [] } ] } ] }, { "id": 2130136, "author_name": "MD. RIPON MIAH", "author_url": "", "post_date": "2023-02-05T07:13:48.877000", "content": "