{
  "id": 568802,
  "title": "Using Local Training and test_sequence for Inference Submission",
  "url": "/competitions/stanford-rna-3d-folding/discussion/568802",
  "author_name": "",
  "post_date": "2025-03-18T03:32:26.803711100Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hi everyone, I have a question regarding the submission process. Is it acceptable to train my model on my local computer and then use the test_sequence for inference? Can I directly upload the resulting predictions as my submission file? Any clarification on this approach would be greatly appreciated. Thank you!</p>",
  "messages": [
    {
      "id": "3152662",
      "postDate": "03/18/2025 03:32:26",
      "content": "<p>Hi everyone, I have a question regarding the submission process. Is it acceptable to train my model on my local computer and then use the test_sequence for inference? Can I directly upload the resulting predictions as my submission file? Any clarification on this approach would be greatly appreciated. Thank you!</p>",
      "rawMarkdown": "Hi everyone, I have a question regarding the submission process. Is it acceptable to train my model on my local computer and then use the test_sequence for inference? Can I directly upload the resulting predictions as my submission file? Any clarification on this approach would be greatly appreciated. Thank you!",
      "votes": null
    },
    {
      "id": "3186827",
      "postDate": "04/25/2025 07:54:22",
      "content": "<p>You can train it locally.  Upload the model to a Kaggle Notebook and do the inference against the test_sequence.csv.  Save the results to submission.csv then submit it</p>",
      "rawMarkdown": "You can train it locally.  Upload the model to a Kaggle Notebook and do the inference against the test_sequence.csv.  Save the results to submission.csv then submit it",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3186827,
      "author_name": "ox17ng",
      "author_url": "",
      "post_date": "04/25/2025 07:54:22",
      "content": "<p>You can train it locally.  Upload the model to a Kaggle Notebook and do the inference against the test_sequence.csv.  Save the results to submission.csv then submit it</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3152662": "Hi everyone, I have a question regarding the submission process. Is it acceptable to train my model on my local computer and then use the test_sequence for inference? Can I directly upload the resulting predictions as my submission file? Any clarification on this approach would be greatly appreciated. Thank you!",
    "3186827": "You can train it locally.  Upload the model to a Kaggle Notebook and do the inference against the test_sequence.csv.  Save the results to submission.csv then submit it"
  },
  "source": "meta"
}