cl886699

loss一直降，但是预测出来的框完全不对

4

你好，我这训练自己的数据，都是小物体，都在30pixel以下，训练集上loss下降很快，第一阶段训练完，loss较低，但是预测没有任何框，第二阶段训练后，loss略有上升，预测的框全是大框，和标注完全对不上 

train_eager.py不能训练了

1

你好，你在改代码吗，train_eager.py好像没有训练代码了

update the gradient do not need freeze one of network?

2

when update the gradient , did not freeze g or d in code ` @tf.function def train_step(lr, hr): with tf.GradientTape(persistent=True) as tape: sr = generator(lr, training=True) hr_output = discriminator(hr, training=True)...

多gpu训练报错

1

你好，我用多gpu训练模型时，报了以下错误，我想要多gpu训练，还需要更改哪些呢 faster_rcnn\detection\utils\misc.py:53 calc_batch_padded_shape * return tf.cast(tf.reduce_max(meta[:, 6:8], axis=0), tf.int32).numpy() AttributeError: 'Tensor' object has no attribute 'numpy'

about training

17

hi, boss, i trained on levircd datasets, when i just overfit one image, all losses could decline quickly and got a pretty results. but when i training on all after...

### Search before asking - [X] I have searched the YOLOv8 [issues](https://github.com/ultralytics/ultralytics/issues) and found no similar bug report. ### YOLOv8 Component _No response_ ### Bug I finetuning on dotav2 datasets...

cuda error: out of memory

3

could you please tell me how many graphic memories are required? can i use my 6gb memories to run this project

I use this command pyinstaller anylabeling.spec when running this .exe，but it can not use gpu. how i can use pyinstaller build a .exe using gpu? thanks

Can't run this demo on hugging face

1


I hope this message finds you well. I have been using the 3dgs library for training and aligning 3D geospatial models in the Earth-Centered, Earth-Fixed (ECEF) coordinate system. While experimenting...