NEMOlv

Is there any way I can reproduce the performance of the model consistently?

1

When I was experimenting with prompt learning, I found that the performance of the model was very unstable, with performance fluctuating from high to low. Is there any way I...

Hi, thank you very much for your open source. I have a few small questions that I hope you can answer. What is the logic for writing the process_logits() function...

Papers Weighted CE Loss and Focal Loss are set to the opposite

1

Papers Weighted CE Loss and Focal Loss are set to the opposite

Suppose I have the following probs and labels (Binary classification)： ``` probs = torch.FloatTensor([[0.3], [0.8], [0.2], [0.7]]) targets = torch.LongTensor([[0], [1], [0], [1]]) ``` Execute the following code: ``` loss...

A question about BinaryDSCLoss

2

Thank you very much for your summary of the loss function in the field of NLP. And, I have a question about BinaryDSCLoss. I sincerely hope you can take time...

Hi, this PR contains a three-part modification. PART 1：Modify the Remove & Close function for segment.go I think the "seg.closed" for segment files should be set to "false" after calling...