AlphaNext comments

Results 19 comments of


                                            AlphaNext

ImportError: pycurl: libcurl link-time version (7.37.1) is older than compile-time version (7.43.0)

you should reinstall curl library, for centos user, you can finish it with this link [Install curl From Source](https://noknow.info/it/os/install_curl_from_source)

'MaskPseudoSampler is already registered in bbox_sampler'

Hi @Ken-97 , I use the following version and solve this error. Hope this could help you. ``` mmcv-full == v1.3.14 mmdetection == v2.17.0 ```

loss变为NAN

Hello, @zhimingluo 有几个问题想请教一下，下面的程序片段摘自NLDF.py中： ``` # 73~84行 #Get the contour term self.Prob_C = tf.reshape(self.Prob, [1, 176, 176, 2]) self.Prob_Grad = tf.tanh(self.im_gradient(self.Prob_C)) self.Prob_Grad = tf.tanh(tf.reduce_sum(self.im_gradient(self.Prob_C), reduction_indices=3, keep_dims=True)) self.label_C = tf.reshape(self.label_holder, [1, 176,...

训练自己的数据集过程中出错

最近同样遇到这样的问题，可能是gt中的txt文件格式不太对。具体可以参考这个链接：[yolov5](https://docs.ultralytics.com/tutorials/train-custom-datasets/ ) ``` The *.txt file specifications are: * One row per object * Each row is class x_center y_center width height format. * Box coordinates must be in normalized...

ModuleNotFoundError: No module named 'fused_layer_norm_cuda'

> 看起来是apex的锅。我找到了解决办法，在apex库的issues中翻到了，[NVIDIA/apex#1737](https://github.com/NVIDIA/apex/issues/1737) 我的ubuntu系统是22.04，所以apex切换一个“22.04-dev”的分支，再安装就行了。用22.04-dev分支还是报错，```enable_layernorm_kernel=False```也不管用。另外直接```# text to video python scripts/inference.py configs/opensora-v1-1/inference/sample.py \ --ckpt-path CKPT_PATH --prompt "A beautiful sunset over the city" --num-frames 32 --image-size 480 854``` 出现如下错误： ``` RuntimeError: Could...

Question consultation

> Dear, Recently I have read your project, and I feel instersted in it, i try to run your project, and see the performance of the model. And during this...

mobilenetv3 代码实现似乎和论文有出入

应该不缺吧，你可以对比下[TensorFlow官方实现](https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet_v3.py#L25-L27)和[PyTorch的官方实现](https://github.com/pytorch/vision/blob/135a0f9ea9841b6324b4fe8974e2543cbb95709a/torchvision/models/mobilenetv3.py#L69-L94)

AlphaNext