So, I used this function from rxrx1-utils to import images in rgb format, while loading the images in rgb format from there, my Kaggle kernel gets crashed everytime, how do I prevent this from happening?\nPlease help!\nAnd yes, my GPU is on.
", "messages": [ { "id": "571181", "postDate": "07/09/2019 09:07:57", "content": "So, I used this function from rxrx1-utils to import images in rgb format, while loading the images in rgb format from there, my Kaggle kernel gets crashed everytime, how do I prevent this from happening?\nPlease help!\nAnd yes, my GPU is on.
", "rawMarkdown": "So, I used this function from [rxrx1-utils](https://github.com/recursionpharma/rxrx1-utils/blob/master/rxrx/io.py) to import images in rgb format, while loading the images in rgb format from there, my Kaggle kernel gets crashed everytime, how do I prevent this from happening?\nPlease help!\nAnd yes, my GPU is on.", "votes": null }, { "id": "571442", "postDate": "07/09/2019 15:39:29", "content": "What's the error?
", "rawMarkdown": "What's the error?", "votes": null }, { "id": "571460", "postDate": "07/09/2019 15:52:17", "content": "I didn't get any error, it just crashed and started again.\nSince, I was importing images in RGB format, I think it couldn't handle this much data and crashed, what do you think?
", "rawMarkdown": "I didn't get any error, it just crashed and started again.\nSince, I was importing images in RGB format, I think it couldn't handle this much data and crashed, what do you think?", "votes": null }, { "id": "571483", "postDate": "07/09/2019 16:34:52", "content": "If you tried to load all of the data at once, that would definitely cause a crash.
", "rawMarkdown": "If you tried to load all of the data at once, that would definitely cause a crash.", "votes": null }, { "id": "571487", "postDate": "07/09/2019 16:44:17", "content": "I tried to load the training data, i.e. 36515*12 images :/
", "rawMarkdown": "I tried to load the training data, i.e. 36515*12 images :/", "votes": null }, { "id": "601656", "postDate": "08/18/2019 00:58:23", "content": "Data loading RAM is too large
", "rawMarkdown": "Data loading RAM is too large", "votes": null }, { "id": "602220", "postDate": "08/18/2019 19:32:45", "content": "Hi @sarques , it's hard to help you debug without seeing your code. Can you make your kernel public so we can help you better? And what is the error message you are getting? Any code example would help here.
", "rawMarkdown": "Hi @sarques , it's hard to help you debug without seeing your code. Can you make your kernel public so we can help you better? And what is the error message you are getting? Any code example would help here.", "votes": null }, { "id": "602221", "postDate": "08/18/2019 19:34:48", "content": "another simple thing you could do is start from a simple working kernel, fork it, and improve on it - like this one https://www.kaggle.com/yhn112/resnet18-baseline-pytorch-ignite
", "rawMarkdown": "another simple thing you could do is start from a simple working kernel, fork it, and improve on it - like this one https://www.kaggle.com/yhn112/resnet18-baseline-pytorch-ignite", "votes": null } ], "comments": [ { "id": 571442, "author_name": "mrosett", "author_url": "", "post_date": "07/09/2019 15:39:29", "content": "What's the error?
", "votes": null, "replies": [ { "id": 571460, "author_name": "sarques", "author_url": "", "post_date": "07/09/2019 15:52:17", "content": "I didn't get any error, it just crashed and started again.\nSince, I was importing images in RGB format, I think it couldn't handle this much data and crashed, what do you think?
", "votes": null, "replies": [] } ] }, { "id": 571483, "author_name": "mrosett", "author_url": "", "post_date": "07/09/2019 16:34:52", "content": "If you tried to load all of the data at once, that would definitely cause a crash.
", "votes": null, "replies": [ { "id": 571487, "author_name": "sarques", "author_url": "", "post_date": "07/09/2019 16:44:17", "content": "I tried to load the training data, i.e. 36515*12 images :/
", "votes": null, "replies": [] } ] }, { "id": 601656, "author_name": "", "author_url": "", "post_date": "08/18/2019 00:58:23", "content": "Data loading RAM is too large
", "votes": null, "replies": [] }, { "id": 602220, "author_name": "michelml", "author_url": "", "post_date": "08/18/2019 19:32:45", "content": "Hi @sarques , it's hard to help you debug without seeing your code. Can you make your kernel public so we can help you better? And what is the error message you are getting? Any code example would help here.
", "votes": null, "replies": [ { "id": 602221, "author_name": "michelml", "author_url": "", "post_date": "08/18/2019 19:34:48", "content": "another simple thing you could do is start from a simple working kernel, fork it, and improve on it - like this one https://www.kaggle.com/yhn112/resnet18-baseline-pytorch-ignite
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "571181": "So, I used this function from [rxrx1-utils](https://github.com/recursionpharma/rxrx1-utils/blob/master/rxrx/io.py) to import images in rgb format, while loading the images in rgb format from there, my Kaggle kernel gets crashed everytime, how do I prevent this from happening?\nPlease help!\nAnd yes, my GPU is on.", "571442": "What's the error?", "571460": "I didn't get any error, it just crashed and started again.\nSince, I was importing images in RGB format, I think it couldn't handle this much data and crashed, what do you think?", "571483": "If you tried to load all of the data at once, that would definitely cause a crash.", "571487": "I tried to load the training data, i.e. 36515*12 images :/", "601656": "Data loading RAM is too large", "602220": "Hi @sarques , it's hard to help you debug without seeing your code. Can you make your kernel public so we can help you better? And what is the error message you are getting? Any code example would help here.", "602221": "another simple thing you could do is start from a simple working kernel, fork it, and improve on it - like this one https://www.kaggle.com/yhn112/resnet18-baseline-pytorch-ignite" }, "source": "meta" }