chenhao
chenhao
0.90几吧，换成bert large 分数能到0.91+，
bert +esim 比只用bert要低一点
@seasonyang i meet the same confusion,don't know where is the model by training .During the training ,the ssd_300_vgg is not changed .
this is the code ： training, validation, test = keras_rcnn.datasets.malaria_phenotypes.load_data() classes = { "rbc": 1, "leu": 2, "ring": 3, "tro": 4, "sch": 5, "gam": 6} generator = keras_rcnn.preprocessing.ObjectDetectionGenerator() generator =...