wenchima

Does anybody figure it out?

I think the above problem is solved by just transforming hsv to array in the script of `yolodet/dataset/pipelines/transforms.py`. `hsv = np.asarray(hsv) `

Now I can train the model but it keeps popping up the error below during the training process. ``` error in modulated_deformable_im2col_cuda: no kernel image is available for execution on...

I have the similar problem. When I enabled the α=0.4, it seems the output AP would be hurt severely. Do you have any idea how to fix this?

Does it support onnx now?

> 基于VOC数据集，训练RefineDet320，100轮后测试map不到0.2，然而loss一直在下降，想请问一下，这是什么情况？ Hi! I came across the same problem. Have you solved it?

hi @albertyou2 Have you solved your problem? I changed the fc7_conv into fc7 and it worked well. But my problem is that my training loss is always be zero. Have...

Sure. Thank you so much, Pau.