how to generate YOLO_small.ckpt file by runing train.py?

[zimurui]

No description provided.

[cenwfohdkue]

it won't generator .ckpt file in new tensorflow version. just load the .meta and .date.

[Youpretty]

it won't generator .ckpt file in new tensorflow version. just load the .meta and .date.
@hainingbaby hi, hainingbaby. I trained the model with my own dataset, and generated 3 files named "yolo-10.data-00000-of-00001", "yolo-10.index" and 'yolo-10.meta'. How can i load the .meta and .data file? The code below can only open one weight file. Please help, tank you!
parser.add_argument('--weights', default="yolo-10.data-00000-of-00001", type=str)

[supermanhuyu]

@Youpretty
you can test it by "python test.py --weight_dir pascal_voc/output/your_outputdir --weights yolo-10"

[ZhouQianang]

The new tensorflow generate three files(.meta; .index; .data-XXX-of-XXX) instead of .ckpt file. And you can use these files like before. Put them in weights:
yolo.meta
yolo.index
yolo.data-0000-of-0001
Now run: python test.py --weights yolo

[EmbugMaker]

I'm new. I had this problem while running "train.py."Has anyone ever encountered this BUG? How to solve this problem？
File "E:/CODEs/Nerual_Net_work_model/yolo_tensorflow-master/train.py", line 90, in train
train_timer.remain(step, self.max_iter))
ValueError: Unknown format code 'f' for object of type 'str'

[ZhouQianang]

I'm new. I had this problem while running "train.py."Has anyone ever encountered this BUG? How to solve this problem？
File "E:/CODEs/Nerual_Net_work_model/yolo_tensorflow-master/train.py", line 90, in train
train_timer.remain(step, self.max_iter))
ValueError: Unknown format code 'f' for object of type 'str'

I changed the code<<

log_str = '''{} Epoch: {}, Step: {}, Learning rate: {},'''
''' Loss: {:5.3f}\nSpeed: {:.3f}s/iter,'''
'''' Load: {:.3f}s/iter, Remain: {}'''.format(
datetime.datetime.now().strftime('%m-%d %H:%M:%S'),
self.data.epoch,
int(step),
round(self.learning_rate.eval(session=self.sess), 6),
loss,
train_timer.average_time,
load_timer.average_time,
train_timer.remain(step, self.max_iter))

to<<

log_str = '''{} Epoch: {}, Step: {}, Learning rate: {},Loss: {:5.3f}\n
Speed: {:.3f}s/iter,Load: {:.3f}s/iter, Remain: {}'''.format(
datetime.datetime.now().strftime('%m-%d %H:%M:%S'),
self.data.epoch,
int(step),
round(self.learning_rate.eval(session=self.sess), 6),
float(loss),
float(train_timer.average_time),
float(load_timer.average_time),
train_timer.remain(step, self.max_iter))