Evaluation script error
Opened this issue · 9 comments
Thank you friends for suggesting interesting things!
python run_metrics.py --models_list weights / eval_list --outputmetric_results.csv
When I do this I get this error:
usage: run_metrics.py [-h] [--models_list MODELS_LIST] [--output OUTPUT] [--model_root MODEL_ROOT] [--eval_root EVAL_ROOT] [--sample_root SAMPLE_ROOT] [--batch_size BATCH_SIZE] [--eval_samples EVAL_SAMPLES] [--device DEVICE] run_metrics.py: error: unrecognized arguments: / eval_list --outputmetric_results.csv
It looks like the arguments aren't working. I want you to find a concrete solution.
thank you!
Can you try removing the space like this?
python run_metrics.py --models_list weights/eval_list --output metric_results.csv
Another error will occur.
`al_list --output metric_results.csv
multiprocessing.pool.RemoteTraceback:
"""
Traceback (most recent call last):
File "/home/test/anaconda3/envs/gan/lib/python3.6/multiprocessing/pool.py", line 119, in worker
result = (True, func(*args, **kwds))
File "/home/test/GANSketching/run_metrics.py", line 105, in run_vgg
from eval.precision_recall import metrics as pr
ImportError: cannot import name 'metrics'
"""
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "run_metrics.py", line 155, in
metrics[name] = get_metrics(opt, name, target)
File "run_metrics.py", line 56, in get_metrics
stats_fake = get_stats(opt, g, fake_folder)
File "run_metrics.py", line 87, in get_stats
vgg_features = get_vgg_features(folder, opt.eval_samples, opt.batch_size)
File "run_metrics.py", line 101, in get_vgg_features
return p.apply(run_vgg, (folder, eval_samples, batch_size,))
File "/home/test/anaconda3/envs/gan/lib/python3.6/multiprocessing/pool.py", line 259, in apply
return self.apply_async(func, args, kwds).get()
File "/home/test/anaconda3/envs/gan/lib/python3.6/multiprocessing/pool.py", line 644, in get
raise self._value
ImportError: cannot import name 'metrics'
`
It could be that you haven't pulled the submodules. I would try running git submodule update --init --recursive in the repo and then run the script.
As instructed, I ran it in the following order, but I get the same error.
①git submodule update --init --recursive
②bash scripts/train_teaser_cat.sh
③Access 'https://wandb.ai' -> The execution result is generated in / GANSketching / wandb.
④python run_metrics.py --models_list weights/eval_list --output metric_results.csv ->Here, the previous error appears
Can you check if metrics.py exists in eval/precision_recall? It should exist if you pulled the submodules
eval/precision_recall was missing metrics.py
Why doesn't git submodule update --init --recursive respond?
I'm not sure about the cause, but as a temporary fix, you can pull the precision_recall directly at here: https://github.com/PeterWang512/precision_recall
and move the folder to in the eval folder
I installed tensorflow-gpu == 1.5.0 because of a dnnlib error.
However, this new error has occurred and I would like to ask for a solution.
`al_list --output metric_results.csv
/home/anaconda3/envs/gan/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:493: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
/home/anaconda3/envs/gan/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:494: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/home/anaconda3/envs/gan/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:495: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
/home/anaconda3/envs/gan/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:496: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/home/anaconda3/envs/gan/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:497: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
/home/anaconda3/envs/gan/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:502: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
multiprocessing.pool.RemoteTraceback:
"""
Traceback (most recent call last):
File "/home/anaconda3/envs/gan/lib/python3.6/multiprocessing/pool.py", line 119, in worker
result = (True, func(*args, **kwds))
File "/home/GANSketching/run_metrics.py", line 105, in run_vgg
from eval.precision_recall import metrics as pr
File "/home/GANSketching/eval/precision_recall/metrics.py", line 13, in
from utils import init_tf
ImportError: cannot import name 'init_tf'
"""
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "run_metrics.py", line 155, in
metrics[name] = get_metrics(opt, name, target)
File "run_metrics.py", line 56, in get_metrics
stats_fake = get_stats(opt, g, fake_folder)
File "run_metrics.py", line 87, in get_stats
vgg_features = get_vgg_features(folder, opt.eval_samples, opt.batch_size)
File "run_metrics.py", line 101, in get_vgg_features
return p.apply(run_vgg, (folder, eval_samples, batch_size,))
File "/home/anaconda3/envs/gan/lib/python3.6/multiprocessing/pool.py", line 259, in apply
return self.apply_async(func, args, kwds).get()
File "/home/anaconda3/envs/gan/lib/python3.6/multiprocessing/pool.py", line 644, in get
raise self._value
ImportError: cannot import name 'init_tf'
`
I'll look into this issue, and see if I can reproduce this error. Stay tuned!