Dear authors,
First and foremost, I deeply appreciate your outstanding research work and the open-source code repository you have shared—your contribution has been incredibly valuable to the research community and my own work.
I am currently attempting to reproduce the experimental results reported in your paper using the released code. However, after numerous attempts with different hyperparameter combinations, I have been unable to replicate the performance stated in the paper. I suspect this challenge stems from the large number of tunable hyperparameters in the codebase, coupled with the lack of explicit, task-specific parameter configurations in the repository or documentation.
To help resolve this reproduction issue and enable broader verification of your impactful work, I sincerely request that you consider publicly releasing:
- Complete experimental results;
- Exact hyperparameter settings corresponding to each task;
- Full log files from the
output directory.
Sharing these materials would not only help me overcome the current reproduction bottleneck but also allow other researchers to validate, build upon, and extend your work—ultimately advancing the entire field.
Thank you very much for your time and consideration. I look forward to your reply at your earliest convenience.
Dear authors,
First and foremost, I deeply appreciate your outstanding research work and the open-source code repository you have shared—your contribution has been incredibly valuable to the research community and my own work.
I am currently attempting to reproduce the experimental results reported in your paper using the released code. However, after numerous attempts with different hyperparameter combinations, I have been unable to replicate the performance stated in the paper. I suspect this challenge stems from the large number of tunable hyperparameters in the codebase, coupled with the lack of explicit, task-specific parameter configurations in the repository or documentation.
To help resolve this reproduction issue and enable broader verification of your impactful work, I sincerely request that you consider publicly releasing:
outputdirectory.Sharing these materials would not only help me overcome the current reproduction bottleneck but also allow other researchers to validate, build upon, and extend your work—ultimately advancing the entire field.
Thank you very much for your time and consideration. I look forward to your reply at your earliest convenience.