Justin
Justin
change # TODO: use binary_crossentropy instead class_loss = obj_mask * sparse_categorical_crossentropy( true_class_idx, pred_class) to true_class_onehot = tf.one_hot(tf.cast(true_class_idx, tf.int64), depth=classes, axis=-1) true_class_binary = tf.reshape(true_class_onehot, (tf.shape(y_true)[0], grid_size,grid_size,tf.shape(y_true)[3],-1, 1)) pred_class_binary = tf.reshape(pred_class, (tf.shape(y_true)[0],...
小图片的结果是图像镜像反转的效果，大图像就没这个问题
 As I said, 3d scatter colorscale should not bind with x3, it should be set independently.I just want a 3d scatter with x1,x2,x3 and color is used to help...
 环境貌似没有问题，是什么原因呢？
目前的kubeflow_dashboard的Dockerfile和base的dockerfile有问题，打镜像过程就会报错，各种源问题和版本问题，打镜像报错只能是dockerfile有问题了，希望更新正确的dockerfile
pocket: warrior of fate rom no sound when using the mra.exe to convert zip to rom I forgot the qsound.zip during my first try. It failed. the second try worked...
女朋友那可以创建任务，我这里任务创建不了也看不到她创建的
代码： ``` from ltp import StnSplit,LTP sents = StnSplit().split("汤姆生病了。他去了医院。") sents = StnSplit().batch_split(["他叫汤姆去拿外衣。", "汤姆生病了。他去了医院。"]) print(sents) ltp = LTP("./LTP_model/base/") cws, pos, ner = ltp.pipeline(["他叫汤姆去拿外衣。"], tasks=["cws", "pos", "ner"]).to_tuple() # to tuple 可以自动转换为元组格式 #...
### Is there an existing issue for this? - [x] I have searched the existing issues ### Describe the bug ```python dbClient = MilvusClient("milvus_solax.db") schema = MilvusClient.create_schema( auto_id=True, enable_dynamic_field=True, )...