Closed rsaite closed 3 years ago
I tried running python reproduce_results.py, subsetting the performed experiments with
python reproduce_results.py
models = [ conf_fastai_xresnet1d101 # , # conf_fastai_resnet1d_wang, # conf_fastai_lstm, # conf_fastai_lstm_bidir, # conf_fastai_fcn_wang, # conf_fastai_inception1d, # conf_wavelet_standard_nn, ] ########################################## # STANDARD SCP EXPERIMENTS ON PTBXL ########################################## experiments = [ # ('exp0', 'all'), ("exp1", "diagnostic") # , # ('exp1.1', 'subdiagnostic'), # ('exp1.1.1', 'superdiagnostic'), # ('exp2', 'form'), # ('exp3', 'rhythm') ]
I observed the following issue:
Training from scratch... model: fastai_xresnet1d101 epoch train_loss valid_loss time 0 3.974695 #na# 00:37 LR Finder is complete, type {learner_name}.recorder.plot() to see the graph. epoch train_loss valid_loss time 0 0.749000 0.617362 00:26 1 0.473438 0.320645 00:26 2 0.176713 0.144030 00:26 3 0.108897 0.087803 00:26 4 0.090648 0.088311 00:26 5 0.087162 1.318479 00:26 6 0.081824 0.075645 00:26 7 0.082379 0.417358 00:26 8 0.076183 0.074853 00:26 9 0.073203 0.072046 00:26 10 0.069833 0.071563 00:26 11 0.068211 0.072181 00:27 12 0.066698 0.068252 00:27 13 0.064080 0.065109 00:27 14 0.062807 0.064614 00:27 15 0.062573 0.064388 00:27 16 0.060904 0.063040 00:27 17 0.060706 0.061243 00:27 18 0.058977 0.061040 00:27 19 0.058891 0.062251 00:27 20 0.058370 0.060759 00:27 21 0.056846 0.059595 00:27 22 0.056774 0.059208 00:27 23 0.055560 0.062082 00:27 24 0.055347 0.059120 00:27 25 0.054211 0.059825 00:27 26 0.054380 0.060731 00:27 27 0.054117 0.061301 00:27 28 0.053299 0.060714 00:27 29 0.052234 0.059627 00:27 30 0.051338 0.058611 00:27 31 0.050798 0.056932 00:27 32 0.050442 0.056777 00:27 33 0.049988 0.060446 00:27 34 0.048720 0.057679 00:26 35 0.048680 0.057815 00:26 36 0.047615 0.056821 00:26 37 0.046539 0.057558 00:26 38 0.046356 0.057474 00:26 39 0.046132 0.057675 00:26 40 0.045098 0.057444 00:26 41 0.044321 0.057479 00:26 42 0.043867 0.057333 00:26 43 0.043355 0.058301 00:27 44 0.043441 0.057717 00:26 45 0.042899 0.058201 00:26 46 0.042278 0.057848 00:26 47 0.042592 0.057989 00:26 48 0.043062 0.058147 00:26 49 0.042745 0.058259 00:26 model: fastai_xresnet1d101 aggregating predictions... model: fastai_xresnet1d101 aggregating predictions... model: fastai_xresnet1d101 aggregating predictions... Traceback (most recent call last): File "reproduce_results.py", line 59, in <module> main() File "reproduce_results.py", line 40, in main e.perform() File "/home/rsaite/medalcare/ecg_ptbxl_benchmarking/code/experiments/scp_experiment.py", line 131, in perform ensemble_train.append(np.load(mpath+'y_train_pred.npy')) File "/home/rsaite/anaconda3/envs/ecg_ptbxl/lib/python3.7/site-packages/numpy/lib/npyio.py", line 457, in load raise ValueError("Cannot load file containing pickled data " ValueError: Cannot load file containing pickled data when allow_pickle=False
After setting allow_pickle=True, the script runs to completion.
allow_pickle=True
I tried running
python reproduce_results.py
, subsetting the performed experiments withI observed the following issue:
After setting
allow_pickle=True
, the script runs to completion.