search

tnikolla / robot-grasp-detection

Detecting robot grasping positions with deep neural networks. The model is trained on Cornell Grasping Dataset. This is an implementation mainly based on the paper 'Real-Time Grasp Detection Using Convolutional Neural Networks' from Redmon and Angelova.
Apache License 2.0
238 stars 84 forks source link

error when run grap_det.py #22

Open dingshenglan opened 6 years ago

dingshenglan commented 6 years ago

when I run grasp_det.py, I met this error as follow: ValueError: Tensor conversion requested dtype int32 for Tensor with dtype float64: 'Tensor("truediv:0", shape=(), dtype=float64, device=/device:CPU:0)' How can I solve this error?????

I had used "build_cgd_dataset.py" to get tfrecords.

I has made small changes in "grasp_det.py" as follow:

TRAIN_FILE = r'F:\RGB-D_grasp_datasets\data01\01\train-cgd' VALIDATE_FILE = r'F:\RGB-D_grasp_datasets\data01\01\validation-cgd'

parser.add_argument( '--data_dir', type=str, default=r'F:\RGB-D_grasp_datasets\data01\01', help='Directory with training data.' ) parser.add_argument( '--log_dir', type=str, default=r'F:\robot-grasp-detection-master', help='Tensorboard log_dir.' parser.add_argument( '--model_path', type=str, default=r'F:\robot-grasp-detection-master\models\grasp\m4.ckpt', help='Variables for the model.' )

lx-onism commented 6 years ago

I met the same problem and had no solution.Have you solved the question? I would like to get some help!!!

jinhuan-hit commented 5 years ago

@dingshenglan @lx-onism You can use them by giving them as arguments,check issue #5,for example: ./imagenet_classifier.py --batch_size=128 --model_path=./models/imagenet/m1/m1.ckpt --train_or_validation=train

le-wei commented 3 years ago

I cannot download the pre-training weights now, can you share a copy with me. thank you very much. lvlvlup@163.com @lx-onism @jinhuan-hit

SpaceHan commented 3 years ago

that's caused by the "/8" operation, I just defined a fixed maxvalue and skipepd the problem

zhoumo1121 commented 3 years ago

ValueError: Tensor conversion requested dtype int32 for Tensor with dtype float64: 'Tensor("truediv:0", shape=(), dtype=float64, device=/device:CPU:0)' May I ask how to solve this problem? Thanks, I've been working on it for a long time!

zhoumo1121 commented 3 years ago

when I run grasp_det.py, I met this error as follow: ValueError: Tensor conversion requested dtype int32 for Tensor with dtype float64: 'Tensor("truediv:0", shape=(), dtype=float64, device=/device:CPU:0)' How can I solve this error?????

I had used "build_cgd_dataset.py" to get tfrecords.

I has made small changes in "grasp_det.py" as follow:

TRAIN_FILE = r'F:\RGB-D_grasp_datasets\data01\01\train-cgd' VALIDATE_FILE = r'F:\RGB-D_grasp_datasets\data01\01\validation-cgd'

parser.add_argument( '--data_dir', type=str, default=r'F:\RGB-D_grasp_datasets\data01\01', help='Directory with training data.' ) parser.add_argument( '--log_dir', type=str, default=r'F:\robot-grasp-detection-master', help='Tensorboard log_dir.' parser.add_argument( '--model_path', type=str, default=r'F:\robot-grasp-detection-master\models\grasp\m4.ckpt', help='Variables for the model.' )

when I run grasp_det.py, I met this error as follow: ValueError: Tensor conversion requested dtype int32 for Tensor with dtype float64: 'Tensor("truediv:0", shape=(), dtype=float64, device=/device:CPU:0)' How can I solve this error?????

I had used "build_cgd_dataset.py" to get tfrecords.

I has made small changes in "grasp_det.py" as follow:

TRAIN_FILE = r'F:\RGB-D_grasp_datasets\data01\01\train-cgd' VALIDATE_FILE = r'F:\RGB-D_grasp_datasets\data01\01\validation-cgd'

parser.add_argument( '--data_dir', type=str, default=r'F:\RGB-D_grasp_datasets\data01\01', help='Directory with training data.' ) parser.add_argument( '--log_dir', type=str, default=r'F:\robot-grasp-detection-master', help='Tensorboard log_dir.' parser.add_argument( '--model_path', type=str, default=r'F:\robot-grasp-detection-master\models\grasp\m4.ckpt', help='Variables for the model.' )

Has the problem been solved?

zhoumo1121 commented 3 years ago

that's caused by the "/8" operation, I just defined a fixed maxvalue and skipepd the problem

Has the problem been solved?

zhoumo1121 commented 3 years ago

I met the same problem and had no solution.Have you solved the question? I would like to get some help!!!

Has the problem been solved?

zhoumo1121 commented 3 years ago

that's caused by the "/8" operation, I just defined a fixed maxvalue and skipepd the problem

Has the problem been solved?

What is your maximum value？

Shaoli7 commented 1 year ago

ValueError: Tensor conversion requested dtype int32 for Tensor with dtype float64: 'Tensor("truediv:0", shape=(), dtype=float64, device=/device:CPU:0)' May I ask how to solve this problem? Thanks, I've been working on it for a long time!

老哥，这个项目你跑通了吗，能加个qq交流下吗？644966953