Greetings,
\nThis is my first competition on kaggle, I can't submit my notebook, I've tried several techniques to reduce the use of CPU and RAM, my model runs in less than 1 minute, but when I submit it gives that error \"Notebook out of memory\".
\nI'd like to get a copy from someone who's already submitted it to see where I'm going wrong.
Greetings,
\nThis is my first competition on kaggle, I can't submit my notebook, I've tried several techniques to reduce the use of CPU and RAM, my model runs in less than 1 minute, but when I submit it gives that error \"Notebook out of memory\".
\nI'd like to get a copy from someone who's already submitted it to see where I'm going wrong.
You may refer my baseline work, it works on this assignment @edmy135
", "rawMarkdown": "You may refer my baseline work, it works on this assignment @edmy135", "votes": null }, { "id": "2475393", "postDate": "10/09/2023 21:24:51", "content": "@edmy135, fell free to take my notebook as an example to submit
", "rawMarkdown": "edmy135, fell free to take my [notebook](https://www.kaggle.com/code/renatoreggiani/zzzs-ensemble-model-cmi) as an example to submit", "votes": null }, { "id": "2481653", "postDate": "10/14/2023 10:26:25", "content": "If you use tensorflow/keras, make sure to import keras.backend as K
\nand use
\nK.clear_session()
\ngc.collect()
\nin a prediction loop. Keras is known to have issues with memory leaks when running such loops.
\nAlso, to better debug, try to do the inference on the training set and see where it goes wrong.
were you able to send the solution ? Or do you need help ?
", "rawMarkdown": "were you able to send the solution ? Or do you need help ?", "votes": null }, { "id": "2499891", "postDate": "10/26/2023 10:34:47", "content": "When you submit your notebook, the file test_series.parquet gets replaced with a much longer one (similar to how long train_series.parquet is). You can debug your code but running your submit code in interactive mode or commit mode using train_series.parquet in place of test_series.parquet.
Thank you all for your cooperation
", "rawMarkdown": "Thank you all for your cooperation", "votes": null }, { "id": "2500383", "postDate": "10/26/2023 16:12:16", "content": "I still can't get it to submit, this is my basic code block, maybe someone can help me find the error
", "rawMarkdown": "I still can't get it to submit, this is my basic code block, maybe someone can help me find the error", "votes": null } ], "comments": [ { "id": 2475261, "author_name": "ravi20076", "author_url": "", "post_date": "10/09/2023 18:21:31", "content": "You may refer my baseline work, it works on this assignment @edmy135
", "votes": null, "replies": [] }, { "id": 2475393, "author_name": "renatoreggiani", "author_url": "", "post_date": "10/09/2023 21:24:51", "content": "@edmy135, fell free to take my notebook as an example to submit
", "votes": null, "replies": [] }, { "id": 2481653, "author_name": "fritzcremer", "author_url": "", "post_date": "10/14/2023 10:26:25", "content": "If you use tensorflow/keras, make sure to import keras.backend as K
\nand use
\nK.clear_session()
\ngc.collect()
\nin a prediction loop. Keras is known to have issues with memory leaks when running such loops.
\nAlso, to better debug, try to do the inference on the training set and see where it goes wrong.
were you able to send the solution ? Or do you need help ?
", "votes": null, "replies": [] }, { "id": 2499891, "author_name": "cdeotte", "author_url": "", "post_date": "10/26/2023 10:34:47", "content": "When you submit your notebook, the file test_series.parquet gets replaced with a much longer one (similar to how long train_series.parquet is). You can debug your code but running your submit code in interactive mode or commit mode using train_series.parquet in place of test_series.parquet.
Thank you all for your cooperation
", "votes": null, "replies": [] }, { "id": 2500383, "author_name": "edmy135", "author_url": "", "post_date": "10/26/2023 16:12:16", "content": "I still can't get it to submit, this is my basic code block, maybe someone can help me find the error
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2475100": "Greetings, \nThis is my first competition on kaggle, I can't submit my notebook, I've tried several techniques to reduce the use of CPU and RAM, my model runs in less than 1 minute, but when I submit it gives that error \"Notebook out of memory\".\nI'd like to get a copy from someone who's already submitted it to see where I'm going wrong.", "2475261": "You may refer my baseline work, it works on this assignment @edmy135", "2475393": "edmy135, fell free to take my [notebook](https://www.kaggle.com/code/renatoreggiani/zzzs-ensemble-model-cmi) as an example to submit", "2481653": "If you use tensorflow/keras, make sure to import keras.backend as K\nand use\nK.clear_session()\ngc.collect()\nin a prediction loop. Keras is known to have issues with memory leaks when running such loops.\nAlso, to better debug, try to do the inference on the training set and see where it goes wrong.", "2499506": "were you able to send the solution ? Or do you need help ?", "2499891": "When you submit your notebook, the file `test_series.parquet` gets replaced with a much longer one (similar to how long `train_series.parquet` is). You can debug your code but running your submit code in interactive mode or commit mode using `train_series.parquet` in place of `test_series.parquet`.", "2500378": "Thank you all for your cooperation", "2500383": "I still can't get it to submit, this is my basic code block, maybe someone can help me find the error" }, "source": "meta" }