{"cells":[{"metadata":{},"cell_type":"markdown","source":"# Save your time, submit without kernel inference\n\nThis kernel just demonstrates the discussion [Reduce submission time to ~10-15 minutes](https://www.kaggle.com/c/lyft-motion-prediction-autonomous-vehicles/discussion/177662) introduced by Peter(@pestipeti).\n\nThe competition test dataset is provided as dataset.\nSo if we have `submission.csv` predicted locally or by other kernel, we can just submit it."},{"metadata":{},"cell_type":"markdown","source":"# Import already calculated submission.csv as Dataset\n\nWe have 2 choices to prepare dataset:\n\n - 1. Upload calculated submission.csv file (maybe from local machine) as Dataset and use it\n - 2. Use other kernel's dataset (it can be either your own kernel, or any other public kernel)\n\nHere I will demonstrate case2, let me use the best performing kernel's result, from [Pytorch Baseline - Inference](https://www.kaggle.com/pestipeti/pytorch-baseline-inference) again by Peter.\n\nWe can choose \"Add Data\" on right side bar and choose \"Kernel Output Files\" to search the kernel and add its kernel's output as Dataset of this kernel's input.\nFor me the most easiest way to find the kernel is to upvote the kernel you want to use, and filter it as \"Your Favorites\" at search botton. Upvote his kernel! :)"},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"#!ls /kaggle/input/lyft-submission-baseline\n# !ls /kaggle/input/a01-1010-full-resnet18-im128-l5kit110-minfuture10\n!ls /kaggle/input/20201124-ensemble-1-testcsv","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Here it is. We just need to copy it to current directory as output."},{"metadata":{"trusted":true},"cell_type":"code","source":"!cp /kaggle/input/20201124-ensemble-1-testcsv/20201124_ensemble_1_test.csv ./submission.csv","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"That's all! Please enjoy submission without waiting long time:)\n\nHowever please note that use this method only for checking LB score, I guess <span style=\"color:red\">final submission should not be done in this way, it must contain inference code.</span>"}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}