Matthew
Matthew
In readme, performance of CoLA task is defined by accuracy. But this task always measured by Matthew correlation. Does it mean the say thing by just calling Matthew_corr as accuracy?...
https://github.com/google-research/albert/blob/a41cf11700c1ed2b7beab0a2649817fa52c8d6e1/classifier_utils.py#L922 in the repo the first one prediction and second one is label, while in official api doc their positions are reversed. Is there any special thinking behind this? https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/metrics/true_positives
I tried multiple times fine-tuning on CoLA ask, but get Matthew_corr keeps to be 0 in evaluation. I have no idea what happened. I checked file conversion part code and...
Have you consider about adding kerberos support? When I using this awesome package, I found that even executor get the key tab file; it don't use it to connect with...
When I am trying to loading data from mongo to spark, I find one wired performance issue in spark UI. It shows that it split into 2 stages the first...