synlp
Reproduction Issue – Results Significantly Lower Than Reported
Hi,
I have attempted to reproduce your model using the code and hyperparameters provided in the repository and paper. However, my results for both stage 1 and stage 2 are significantly lower than the values reported in your paper. Despite strictly following the given settings, I have been unable to achieve similar performance.
To ensure the reproducibility of the reported results, would you be willing to release the pretrained weights? This would help verify the reliability of the reported values.
hi @LiGuo12 , could you please provide the link for pre trained weights?
Hi,
I have attempted to reproduce your model using the code and hyperparameters provided in the repository and paper. However, my results for both stage 1 and stage 2 are significantly lower than the values reported in your paper. Despite strictly following the given settings, I have been unable to achieve similar performance.
To ensure the reproducibility of the reported results, would you be willing to release the pretrained weights? This would help verify the reliability of the reported values.
Hello, may I ask if you can provide Public Medical Data? I don't seem to have found it in the manuscript.
Hi, I have attempted to reproduce your model using the code and hyperparameters provided in the repository and paper. However, my results for both stage 1 and stage 2 are significantly lower than the values reported in your paper. Despite strictly following the given settings, I have been unable to achieve similar performance. To ensure the reproducibility of the reported results, would you be willing to release the pretrained weights? This would help verify the reliability of the reported values.
Hello, may I ask if you can provide Public Medical Data? I don't seem to have found it in the manuscript.
Your are right, the authors did not provide the public medical data. I generated it by myself from Pubmed. My results are much lower than the values in the manuscript, so I have abandoned this approach.