idealboy
idealboy
**Is it fittable for multi-label train in resnet_model_fn in train.py?** predictions = { 'classes': tf.argmax(logits, axis=1), 'probabilities': tf.nn.softmax(logits, name='softmax_tensor') } accuracy = tf.metrics.accuracy( tf.argmax(labels, axis=1), predictions['classes']) Thank you very much!
> @idealboy For the first question, in the provided code, we have removed the setting of multi-gpus. We may provide it in a close future. > > We modify the...
> > @idealboy For the first question, in the provided code, we have removed the setting of multi-gpus. We may provide it in a close future. > > We modify...
when batchnorm is using, the relu is tf.maximum(alpha*conv, conv), but when batchnorm is false, where is the relu? thank you
yes, I think cudnn-5.1 will be right, thank you!
the dataset.py will load all the data in *.h5 into memory, so you must make sure that the memory is sufficient. And then, your system may be disturbed by a...
重试次数比较多，下载下来的很少
> I am not able to train the model. The command used: > > `python carn/train.py --patch_size 64 --batch_size 64 --max_steps 600000 --decay 400000 --model carn --ckpt_name carn --ckpt_dir checkpoint/carn`...
> I test this code on my dataset，and the result worse than I3D. I guess the main reason is no pre-train model. Yes, I think it converge very slow, and...