KeyError while training for tensorpack

[KD1994]

Hi,

I'm trying to run this on Google colab, while i'm facing this issue for the tensorpack.
Do you have any idea how i can resolve this?

---

KeyError Traceback (most recent call last)
in ()
----> 1 from lib.core.base_trainer.net_work import trainner
2 import setproctitle
3
4 setproctitle.setproctitle("detect")
5

4 frames
/content/drive/My Drive/DSFD-tensorflow-master/lib/core/base_trainer/net_work.py in ()
10 from functools import partial
11
---> 12 from lib.dataset.dataietr import DsfdDataIter
13 from lib.core.model.net.ssd import SSD
14 from train_config import config as cfg

/content/drive/My Drive/DSFD-tensorflow-master/lib/dataset/dataietr.py in ()
10
11 from lib.helper.logger import logger
---> 12 from tensorpack.dataflow import DataFromList
13 from tensorpack.dataflow import BatchData, MultiThreadMapData, MultiProcessPrefetchData
14

/usr/local/lib/python3.6/dist-packages/tensorpack/init.py in ()
6
7 from tensorpack.utils import *
----> 8 from tensorpack.dataflow import *
9
10 # dataflow can be used alone without installing tensorflow

/usr/local/lib/python3.6/dist-packages/tensorpack/dataflow/init.py in ()
46 if not module_name.startswith('_') and
47 module_name not in __SKIP:
---> 48 _global_import(module_name)
49
50

/usr/local/lib/python3.6/dist-packages/tensorpack/dataflow/init.py in _global_import(name)
30 lst = p.all if 'all' in dir(p) else dir(p)
31 if lst:
---> 32 del globals()[name]
33 for k in lst:
34 if not k.startswith('__'):

KeyError: 'base'

Any suggestion would be great.
Thank you

[610265158]

Hi,

I'm trying to run this on Google colab, while i'm facing this issue for the tensorpack.
Do you have any idea how i can resolve this?

KeyError Traceback (most recent call last)
in ()
----> 1 from lib.core.base_trainer.net_work import trainner
2 import setproctitle
3
4 setproctitle.setproctitle("detect")
5
4 frames
/content/drive/My Drive/DSFD-tensorflow-master/lib/core/base_trainer/net_work.py in ()
10 from functools import partial
11
---> 12 from lib.dataset.dataietr import DsfdDataIter
13 from lib.core.model.net.ssd import SSD
14 from train_config import config as cfg
/content/drive/My Drive/DSFD-tensorflow-master/lib/dataset/dataietr.py in ()
10
11 from lib.helper.logger import logger
---> 12 from tensorpack.dataflow import DataFromList
13 from tensorpack.dataflow import BatchData, MultiThreadMapData, MultiProcessPrefetchData
14
/usr/local/lib/python3.6/dist-packages/tensorpack/init.py in ()
6
7 from tensorpack.utils import *
----> 8 from tensorpack.dataflow import *
9
10 # dataflow can be used alone without installing tensorflow
/usr/local/lib/python3.6/dist-packages/tensorpack/dataflow/init.py in ()
46 if not module_name.startswith('_') and
47 module_name not in SKIP:
---> 48 _global_import(module_name)
49
50
/usr/local/lib/python3.6/dist-packages/tensorpack/dataflow/init.py in _global_import(name)
30 lst = p.all if 'all' in dir(p) else dir(p)
31 if lst:
---> 32 del globals()[name]
33 for k in lst:
34 if not k.startswith(''):
KeyError: 'base'

Any suggestion would be great.
Thank you

hi,
Firstly, i sugget that using anaconda, :)
The tensorpack version i use is 0.9.6.
And seems like, it is a python package import problem. U could print the global(), see if there is base.

thaks

[KD1994]

Thanks for the response.
I don't have GPU with me now, so was hoping to train the model on Google colab:sweat_smile:

Anyways thanks for the suggestion I'll take a look into this.