Submissions ends in 2 minutes, scoring 0.0, tried forking many kernels, have done testing on those using my pipeline on train_images path to predict, but submissions does not give me score.
\n\nAny guess?, Feel free to comment and ask for more details.
\n\nThank You!
", "messages": [ { "id": 832153, "postDate": "2020-05-03T22:47:48.303Z", "content": "Submissions ends in 2 minutes, scoring 0.0, tried forking many kernels, have done testing on those using my pipeline on train_images path to predict, but submissions does not give me score.
\n\nAny guess?, Feel free to comment and ask for more details.
\n\nThank You!
", "rawMarkdown": "Submissions ends in 2 minutes, scoring 0.0, tried forking many kernels, have done testing on those using my pipeline on train_images path to predict, but submissions does not give me score.\n\nAny guess?, Feel free to comment and ask for more details.\n\nThank You!", "votes": 1 }, { "id": 846986, "postDate": "2020-05-14T05:34:55.767Z", "content": "@harshitsheoran same here.. did u find any solution ?
", "rawMarkdown": "@harshitsheoran same here.. did u find any solution ?", "replies": [ { "id": 847047, "postDate": "2020-05-14T06:27:05.133Z", "content": "Nah, I just left the competition it pissed me off.
", "rawMarkdown": "Nah, I just left the competition it pissed me off." }, { "id": 847248, "postDate": "2020-05-14T08:57:13.893Z", "content": "what i think is \nwhen u do\nsub_df=sample\nIf \n for dl \n exception
\n\nsub df remains same as sample so you get zero score
", "rawMarkdown": "what i think is \nwhen u do\nsub_df=sample\nIf \n for dl \n exception\n\n \n\nsub df remains same as sample so you get zero score\n ", "votes": 1 }, { "id": 847253, "postDate": "2020-05-14T09:01:16.783Z", "content": "I am gonna try right this moment
", "rawMarkdown": "I am gonna try right this moment" }, { "id": 847271, "postDate": "2020-05-14T09:12:53.120Z", "content": "its running for 3 minutes, this might be it
", "rawMarkdown": "its running for 3 minutes, this might be it" }, { "id": 847409, "postDate": "2020-05-14T11:33:24.710Z", "content": "Submission ran for around 2 hours but the score still remains to be 0.0
", "rawMarkdown": "Submission ran for around 2 hours but the score still remains to be 0.0" }, { "id": 847664, "postDate": "2020-05-14T14:44:32.603Z", "content": "what did u do so it ran in for 2 hrs \nhave u tried to add 0 class or prediction in try exception block ? \nReason i told it is an exception that is causing is because i used different df for sample and predictions. my issued changed from 0.0 to CSV not found which means from the Dataloader loop itself an exception occured so prediction df got never created as csv .hence no csv found\nI suspect exception could be because of no tiles selected for an odd file . Are you iafoss method of Tiling or OpenSlide ?
", "rawMarkdown": "what did u do so it ran in for 2 hrs \nhave u tried to add 0 class or prediction in try exception block ? \nReason i told it is an exception that is causing is because i used different df for sample and predictions. my issued changed from 0.0 to CSV not found which means from the Dataloader loop itself an exception occured so prediction df got never created as csv .hence no csv found\nI suspect exception could be because of no tiles selected for an odd file . Are you iafoss method of Tiling or OpenSlide ?" }, { "id": 848382, "postDate": "2020-05-14T23:25:39.973Z", "content": "Yes, I did append class 0 on exception, I am using my own method of creating data so I take an image, create data from it, predict it and then get the next image.
", "rawMarkdown": "Yes, I did append class 0 on exception, I am using my own method of creating data so I take an image, create data from it, predict it and then get the next image." }, { "id": 848563, "postDate": "2020-05-15T04:31:27.427Z", "content": "I was using open slide . If you using same then there is one kernel using same and ran successfully. It does pre process first using parallel worker and then make prediction . Bit it makes use of level 3 images which are lowest in resolution so will not produce many tiles compared to level 2 ,so m in doubt if we will be crossing the file size or file number limit when we use level 2 images
", "rawMarkdown": "I was using open slide . If you using same then there is one kernel using same and ran successfully. It does pre process first using parallel worker and then make prediction . Bit it makes use of level 3 images which are lowest in resolution so will not produce many tiles compared to level 2 ,so m in doubt if we will be crossing the file size or file number limit when we use level 2 images " }, { "id": 848591, "postDate": "2020-05-15T05:09:21.040Z", "content": "I get below dl error fir train set so some thing similar u too should be getting in dl so it may be abruptly ending the inference loop ,result is zero score
\n\n\nReturn [default_collate(samples) for samples in transposed] File \"/opt/conda/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py\", line 55, in default_collate\n return torch.stack(batch, 0, out=out)\nRuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 0. Got 14 and 6 in dimension 1 at /opt/conda/conda-bld/pytorch_1579022060824/work/aten/src/TH/generic/THTensor.cpp:612\n
I am able to run my submissions without an error, did fix all my problems yesterday, btw I read images from imageio imread
", "rawMarkdown": "I am able to run my submissions without an error, did fix all my problems yesterday, btw I read images from imageio imread" }, { "id": 848659, "postDate": "2020-05-15T06:27:01.033Z", "content": "good to hear that .. what could be causing dataloader issue .. if you can give some input based on error message above..
", "rawMarkdown": "good to hear that .. what could be causing dataloader issue .. if you can give some input based on error message above.." }, { "id": 848762, "postDate": "2020-05-15T08:01:00.170Z", "content": "I am not yet using pytorch on this comp so no idea.
", "rawMarkdown": "I am not yet using pytorch on this comp so no idea." } ] } ], "comments": [ { "id": 846986, "author_name": "Jaideep", "author_url": "", "post_date": "2020-05-14T05:34:55.767000", "content": "@harshitsheoran same here.. did u find any solution ?
", "votes": 0, "replies": [ { "id": 847047, "author_name": "Harshit Sheoran", "author_url": "", "post_date": "2020-05-14T06:27:05.133000", "content": "Nah, I just left the competition it pissed me off.
", "votes": 0, "replies": [] }, { "id": 847248, "author_name": "Jaideep", "author_url": "", "post_date": "2020-05-14T08:57:13.893000", "content": "what i think is \nwhen u do\nsub_df=sample\nIf \n for dl \n exception
\n\nsub df remains same as sample so you get zero score
", "votes": 1, "replies": [] }, { "id": 847253, "author_name": "Harshit Sheoran", "author_url": "", "post_date": "2020-05-14T09:01:16.783000", "content": "I am gonna try right this moment
", "votes": 0, "replies": [] }, { "id": 847271, "author_name": "Harshit Sheoran", "author_url": "", "post_date": "2020-05-14T09:12:53.120000", "content": "its running for 3 minutes, this might be it
", "votes": 0, "replies": [] }, { "id": 847409, "author_name": "Harshit Sheoran", "author_url": "", "post_date": "2020-05-14T11:33:24.710000", "content": "Submission ran for around 2 hours but the score still remains to be 0.0
", "votes": 0, "replies": [] }, { "id": 847664, "author_name": "Jaideep", "author_url": "", "post_date": "2020-05-14T14:44:32.603000", "content": "what did u do so it ran in for 2 hrs \nhave u tried to add 0 class or prediction in try exception block ? \nReason i told it is an exception that is causing is because i used different df for sample and predictions. my issued changed from 0.0 to CSV not found which means from the Dataloader loop itself an exception occured so prediction df got never created as csv .hence no csv found\nI suspect exception could be because of no tiles selected for an odd file . Are you iafoss method of Tiling or OpenSlide ?
", "votes": 0, "replies": [] }, { "id": 848382, "author_name": "Harshit Sheoran", "author_url": "", "post_date": "2020-05-14T23:25:39.973000", "content": "Yes, I did append class 0 on exception, I am using my own method of creating data so I take an image, create data from it, predict it and then get the next image.
", "votes": 0, "replies": [] }, { "id": 848563, "author_name": "Jaideep", "author_url": "", "post_date": "2020-05-15T04:31:27.427000", "content": "I was using open slide . If you using same then there is one kernel using same and ran successfully. It does pre process first using parallel worker and then make prediction . Bit it makes use of level 3 images which are lowest in resolution so will not produce many tiles compared to level 2 ,so m in doubt if we will be crossing the file size or file number limit when we use level 2 images
", "votes": 0, "replies": [] }, { "id": 848591, "author_name": "Jaideep", "author_url": "", "post_date": "2020-05-15T05:09:21.040000", "content": "I get below dl error fir train set so some thing similar u too should be getting in dl so it may be abruptly ending the inference loop ,result is zero score
\n\n\nReturn [default_collate(samples) for samples in transposed] File \"/opt/conda/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py\", line 55, in default_collate\n return torch.stack(batch, 0, out=out)\nRuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 0. Got 14 and 6 in dimension 1 at /opt/conda/conda-bld/pytorch_1579022060824/work/aten/src/TH/generic/THTensor.cpp:612\n
I am able to run my submissions without an error, did fix all my problems yesterday, btw I read images from imageio imread
", "votes": 0, "replies": [] }, { "id": 848659, "author_name": "Jaideep", "author_url": "", "post_date": "2020-05-15T06:27:01.033000", "content": "good to hear that .. what could be causing dataloader issue .. if you can give some input based on error message above..
", "votes": 0, "replies": [] }, { "id": 848762, "author_name": "Harshit Sheoran", "author_url": "", "post_date": "2020-05-15T08:01:00.170000", "content": "I am not yet using pytorch on this comp so no idea.
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "832153": "Submissions ends in 2 minutes, scoring 0.0, tried forking many kernels, have done testing on those using my pipeline on train_images path to predict, but submissions does not give me score.\n\nAny guess?, Feel free to comment and ask for more details.\n\nThank You!", "846986": "@harshitsheoran same here.. did u find any solution ?" } }