wuzifan

Does the python version node still available?

2

I want to learn the python version pbft and change your docker-composer.yml, from node: extends: file: code/node_java/docker-compose.yml service: node_java volumes: - /var/run/docker.sock:/var/run/docker.sock environment: NODE_DEBUGGER: "False" NODE_DEBUG: "False" # API_HOST: "http://192.168.0.42"...

当我按照教程去跑multi gpus的服务时，发现我的服务能够很好的在gpu 1，2，3上启动，但是无法在gpu 0 上进行启动。 调研后，发现在 managed_model.py 这个文件中，有一个设置cuda_visiable_device的方法： ``` python @staticmethod def set_gpu_id(gpu_id=None): if gpu_id: os.environ["CUDA_VISIBLE_DEVICES"] = str(gpu_id) ``` 注意这里的if 的判断条件，当gpu_id =0 时，也会被认为条件失败，从而不设置cuda； 因此，需要手动将其修改为 if gpu_id is not None.

边框回归loss的计算问题

1

在边框回归中，我看到您的代码是这样 ``` PYTHON no_object_loss = tf.reduce_mean(tf.square((1 - y_true[:, 0]) * y_pred[:, 0])) object_loss = tf.reduce_mean(tf.square((y_true[:, 0]) * (y_pred[:, 0] - 1))) loss = (tf.reduce_mean(y_true[:, 0] * ( tf.reduce_sum(tf.square(y_true[:, 1:5] -...

[BUG]: Cant Do Validation During Training

1

### 🐛 Describe the bug I try to do validation step on every n steps. the training step goes well, but got this error here is my config: ### Environment...

'DUTIR' 情绪词典脏了

1

` import cntext as ct res = ct.load_pkl_dict('DUTIR.pkl') for k,v in res['DUTIR'].items(): print(k,len(v)) print("开心" in v) ` 会发现 ’开心‘同时在’乐‘和’惧‘两个类别中出现