tensorflow-yolov3

tensorflow-yolov3 copied to clipboard

I have been trying to understand the LR scheduler. Code can be found [here](https://github.com/YunYang1994/tensorflow-yolov3/blob/03cb272af2e26d598c553f3a2d38024fc6f67a0b/train.py#L64) Explanation with graph would be much appreciated

harshdhamecha

victor30608

Converted `bbox_class_ind` to integer data type to get rid of IndexError when bboxes is of float type.

harshdhamecha

运行 train.py 进行到Restoring weights时segmentation fault

4

`python3 train.py` `2019-08-05 06:48:09.794696: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA` `2019-08-05 06:48:10.439188: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:964] successful NUMA node read from...

TDisWatchingU

cpu训练正常，gpu训练出现loss=nan

1

我训练的是自制的数据集，1000×500的tif图片 只更改了config.py中的图片和标签的dateset.txt和clsss.names 

DragMou1204

Hello. I get the following error when running evaluate.py. Unfortunately, I have not found a solution to my problem. Help me please Error : NotFoundError (see above for traceback): Restoring...

Amirhosein-mkh

Bumps [tensorflow-gpu](https://github.com/tensorflow/tensorflow) from 1.11.0 to 2.12.0. Release notes Sourced from tensorflow-gpu's releases. TensorFlow 2.12.0 Release 2.12.0 TensorFlow Breaking Changes Build, Compilation and Packaging Removed redundant packages tensorflow-gpu and tf-nightly-gpu. These...

dependabot[bot]

关于分布训练loss值

7

貌似存在过拟合的问题，训练70个周期，Test loss比 Trainloss大很多，只训练20个周期，两者数值差不多，请问有什么解决方法么 Epoch: 70 Time: 2019-09-27 07:05:06 Train loss: 2.58 Test loss: 9.61 Saving ./checkpoint/yolov3_test_loss=9.6080.ckpt ...

navy63

如何将accuracy在tensorboard上面显示出来啊？

4

修改这个代码的地方我找到了，但是后面不知道怎么改了 with tf.name_scope('summary'): tf.summary.scalar("learn_rate", self.learn_rate) tf.summary.scalar("giou_loss", self.giou_loss) tf.summary.scalar("conf_loss", self.conf_loss) tf.summary.scalar("prob_loss", self.prob_loss) tf.summary.scalar("total_loss", self.loss) #tf.summary.scalar("accuracy", 这里不知道怎么写了) logdir = "./data/log/" if os.path.exists(logdir): shutil.rmtree(logdir) os.mkdir(logdir) self.write_op = tf.summary.merge_all() self.summary_writer = tf.summary.FileWriter(logdir, graph=self.sess.graph)

wl4135

tensorRT不支持tensorflow的内置函数

6

用tensorRT进行模型加速，但是tensorRT不支持tensorflow的tf.cast，tf.tile，tf.range等函数，这几个函数在yolov3.py里，不知道用什么方法改成tensorRT能识别的函数

hongtao-ai