Datasets:

ibrahimhamamci

/

CT-RATE

like 43

Hi @Henry2001 , It actually works in my case except for volumes, which I believe is because the timeout occurs when glob takes too long with too many .nii.gz images.

The error for the volumes configuration is:

{
  "error": "An error occurred while generating the dataset",
  "cause_exception": "DatasetGenerationError",
  "cause_message": "An error occurred while generating the dataset",
  "cause_traceback": [
    "Traceback (most recent call last):\n",
    " File \"/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py\", line 2005, in _prepare_split_single\n num_examples, num_bytes = writer.finalize()\n",
    " File \"/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py\", line 610, in finalize\n raise SchemaInferenceError(\"Please pass `features` or at least one example when writing data\")\n",
    "datasets.arrow_writer.SchemaInferenceError: Please pass `features` or at least one example when writing data\n",
    "\nThe above exception was the direct cause of the following exception:\n\n",
    "Traceback (most recent call last):\n",
    " File \"/src/services/worker/src/worker/job_manager.py\", line 125, in process\n job_result = self.job_runner.compute()\n",
    " File \"/src/services/worker/src/worker/job_runners/config/parquet_and_info.py\", line 1398, in compute\n compute_config_parquet_and_info_response(\n",
    " File \"/src/services/worker/src/worker/job_runners/config/parquet_and_info.py\", line 1306, in compute_config_parquet_and_info_response\n parquet_operations, partial = stream_convert_to_parquet(\n",
    " File \"/src/services/worker/src/worker/job_runners/config/parquet_and_info.py\", line 901, in stream_convert_to_parquet\n builder._prepare_split(\n",
    " File \"/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py\", line 1860, in _prepare_split\n for job_id, done, content in self._prepare_split_single(\n",
    " File \"/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py\", line 2016, in _prepare_split_single\n raise DatasetGenerationError(\"An error occurred while generating the dataset\") from e\n",
    "datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset\n"
  ]
}

I assume it's because we don't support the NIFTI format (.nii) at the moment, see https://github.com/huggingface/datasets/issues/3583#issuecomment-1038284498.

So, feel free to chime in the issue above or upvote; also: maybe the volumes config should be removed for now, as we cannot process it.

cc @mariosasko @lhoestq @albertvillanova @polinaeterna fyi

Hi @severo , thank you very much! I will remove volumes for now.