\n
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3852696%2F978db01581299702267c1b0cbceb2e9f%2Ftemp1.png?generation=1680345375034649&alt=media\")
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3852696%2F5825c49b86abdf06e7002405fcfdc1aa%2Ftemp2.png?generation=1680345378848587&alt=media\")
This score is mean to the result was producted randomly,what happend?
", "messages": [ { "id": 2872940, "postDate": "2024-06-15T07:30:18.447Z", "content": "Facing similar issue after passing the test data through the trained model on my private Jupyter notebook…, it has to be something related with submission process!
", "rawMarkdown": "Facing similar issue after passing the test data through the trained model on my private Jupyter notebook..., it has to be something related with submission process!", "votes": 1 }, { "id": 2846828, "postDate": "2024-05-31T09:27:23.723Z", "content": "The mapping between id and label is inconsistent.
\nlabels_df = pd.read_csv('D:/data/CIFAR-10 - Object Recognition in Images/trainLabels.csv')
\nlabel_to_int = {label: idx for idx, label in enumerate(labels_df['label'].unique())}
\nint_to_label = {idx: label for label, idx in label_to_int.items()}
print(\"label_to_int:\", label_to_int)
\nprint(\"int_to_label:\", int_to_label)
This score is mean to the result was producted randomly,what happend?
", "rawMarkdown": "This score is mean to the result was producted randomly,what happend?", "votes": 17 }, { "id": 1714371, "postDate": "2022-03-06T22:26:14.240Z", "content": "I encountered a similar problem. If you have figured why this happened, please let me know! Much appreciated.
", "rawMarkdown": "I encountered a similar problem. If you have figured why this happened, please let me know! Much appreciated.", "votes": 1 }, { "id": 1569801, "postDate": "2021-11-03T18:06:48.107Z", "content": "Looks to be your model is over fitted on training data, have you done cross validation @dakaipan ?
", "rawMarkdown": "Looks to be your model is over fitted on training data, have you done cross validation @dakaipan ?", "votes": 1 }, { "id": 2964071, "postDate": "2024-08-19T14:00:45.367Z", "content": "It's over fitting, for fix that, you can normalise the dataset and get a validation 10% (with cross validation or other…)
", "rawMarkdown": "It's over fitting, for fix that, you can normalise the dataset and get a validation 10% (with cross validation or other...)" }, { "id": 2459939, "postDate": "2023-09-28T13:33:09.943Z", "content": "Were you able to fix this? How did you do so?
", "rawMarkdown": "Were you able to fix this? How did you do so?" }, { "id": 2412214, "postDate": "2023-08-28T06:38:50.360Z", "content": "it might not be overfitting, which cannot bring such huge gap in most cases.
", "rawMarkdown": "it might not be overfitting, which cannot bring such huge gap in most cases." }, { "id": 2324886, "postDate": "2023-07-01T01:11:00.157Z", "content": "Usually is because overfitting, but since this huge gap, I think you must have bug in generating the final result csv.
", "rawMarkdown": "Usually is because overfitting, but since this huge gap, I think you must have bug in generating the final result csv." }, { "id": 2243390, "postDate": "2023-05-02T21:48:35.383Z", "content": "Possibly it is an error in the prediction sequence. Can you give more details?
", "rawMarkdown": "Possibly it is an error in the prediction sequence. Can you give more details?" }, { "id": 2205200, "postDate": "2023-04-01T10:38:23.710Z", "content": "Have you solved the issue? It's very likely your submission data sort id as string (left), rather than int (right)
\n
It may occurs when your train dataset and test dataset labels are not same. Maybe truck class's label is 1 for your train dataset but it is 2 for yout test dataset. Check their labels.
", "rawMarkdown": "It may occurs when your train dataset and test dataset labels are not same. Maybe truck class's label is 1 for your train dataset but it is 2 for yout test dataset. Check their labels. " }, { "id": 1919285, "postDate": "2022-08-30T09:40:14.420Z", "content": "Is the sequence error in prediction?
", "rawMarkdown": "Is the sequence error in prediction?" }, { "id": 1893449, "postDate": "2022-08-10T20:22:16.157Z", "content": "Did you normalize the feature (e.g., dividing the tensor by 255)?
", "rawMarkdown": "Did you normalize the feature (e.g., dividing the tensor by 255)?" }, { "id": 1872675, "postDate": "2022-07-27T06:54:41.990Z", "content": "Make sure if you properly assigned the labels to the files - I had this problem :D
\n\"listdir(path)\" did not sort files
Facing similar issue after passing the test data through the trained model on my private Jupyter notebook…, it has to be something related with submission process!
", "votes": 1, "replies": [] }, { "id": 2846828, "author_name": "Lee", "author_url": "", "post_date": "2024-05-31T09:27:23.723000", "content": "The mapping between id and label is inconsistent.
\nlabels_df = pd.read_csv('D:/data/CIFAR-10 - Object Recognition in Images/trainLabels.csv')
\nlabel_to_int = {label: idx for idx, label in enumerate(labels_df['label'].unique())}
\nint_to_label = {idx: label for label, idx in label_to_int.items()}
print(\"label_to_int:\", label_to_int)
\nprint(\"int_to_label:\", int_to_label)
I encountered a similar problem. If you have figured why this happened, please let me know! Much appreciated.
", "votes": 1, "replies": [] }, { "id": 1569801, "author_name": "Old Monk", "author_url": "", "post_date": "2021-11-03T18:06:48.107000", "content": "Looks to be your model is over fitted on training data, have you done cross validation @dakaipan ?
", "votes": 1, "replies": [] }, { "id": 2964071, "author_name": "OUARAS Khelil Rafik", "author_url": "", "post_date": "2024-08-19T14:00:45.367000", "content": "It's over fitting, for fix that, you can normalise the dataset and get a validation 10% (with cross validation or other…)
", "votes": 0, "replies": [] }, { "id": 2459939, "author_name": "Dhruv G", "author_url": "", "post_date": "2023-09-28T13:33:09.943000", "content": "Were you able to fix this? How did you do so?
", "votes": 0, "replies": [] }, { "id": 2412214, "author_name": "Pengfei Zhou", "author_url": "", "post_date": "2023-08-28T06:38:50.360000", "content": "it might not be overfitting, which cannot bring such huge gap in most cases.
", "votes": 0, "replies": [] }, { "id": 2324886, "author_name": "XiaosongWen", "author_url": "", "post_date": "2023-07-01T01:11:00.157000", "content": "Usually is because overfitting, but since this huge gap, I think you must have bug in generating the final result csv.
", "votes": 0, "replies": [] }, { "id": 2243390, "author_name": "Jose Torronteras", "author_url": "", "post_date": "2023-05-02T21:48:35.383000", "content": "Possibly it is an error in the prediction sequence. Can you give more details?
", "votes": 0, "replies": [] }, { "id": 2205200, "author_name": "WangHe", "author_url": "", "post_date": "2023-04-01T10:38:23.710000", "content": "Have you solved the issue? It's very likely your submission data sort id as string (left), rather than int (right)
\n
It may occurs when your train dataset and test dataset labels are not same. Maybe truck class's label is 1 for your train dataset but it is 2 for yout test dataset. Check their labels.
", "votes": 0, "replies": [] }, { "id": 1919285, "author_name": "Jack77T", "author_url": "", "post_date": "2022-08-30T09:40:14.420000", "content": "Is the sequence error in prediction?
", "votes": 0, "replies": [] }, { "id": 1893449, "author_name": "shunting", "author_url": "", "post_date": "2022-08-10T20:22:16.157000", "content": "Did you normalize the feature (e.g., dividing the tensor by 255)?
", "votes": 0, "replies": [] }, { "id": 1872675, "author_name": "Jan Sowa", "author_url": "", "post_date": "2022-07-27T06:54:41.990000", "content": "Make sure if you properly assigned the labels to the files - I had this problem :D
\n\"listdir(path)\" did not sort files