Fine-Tuning Job Stuck at “validating_files” for 2 Days

I recently uploaded around 640054 training records for fine-tuning (OpenAI 4.1 nano model). After starting the training job, the status has been stuck at “validating_files” for almost two days.

  • The files were successfully uploaded.

  • The fine-tuning job was created.

  • But it has not moved past the “validating_files” stage.

  • No error message has appeared.

Has anyone else experienced this long delay?
Is this normal for large datasets, or could something be wrong with my upload or dataset structure?