def main(_): # hyper_param_list = def_hyper_param() hyper_param_list = [{'layer': 2, 'feat': [32, 64]}] # {'layer': 1, 'feat': [128]}, # {'layer': 2, 'feat': [128, 8]}, # {'layer': 2, 'feat': [128, 16]}, # {'layer': 2, 'feat': [128, 32]}, # # hyper_param_list = [{'layer': 3, 'feat': [4, 4, 512]}, # {'layer': 3, 'feat': [8, 8, 512]}, # {'layer': 3, 'feat': [4, 4, 256]}, # {'layer': 4, 'feat': [4, 4, 4, 512]}, # {'layer': 4, 'feat': [8, 8, 8, 512]}, # {'layer': 4, 'feat': [4, 4, 4, 256]}, # {'layer': 5, 'feat': [4, 4, 4, 4, 512]}, # {'layer': 5, 'feat': [8, 8, 8, 8, 512]}, # {'layer': 5, 'feat': [4, 4, 4, 4, 128]}] models = [1] for model in models: for hyper_param in hyper_param_list: print("Currently running model: "+str(model)) print("FeatMap: ") print(hyper_param['feat']) # for idx in range(3, len(roi_property.DAT_TYPE_STR)): for idx in range(0, 5): print("Data: " + roi_property.DAT_TYPE_STR[idx]) for subIdx in range(4, 5): print("Subject: " + str(subIdx+1)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat'], name_idx=idx, sub_idx=subIdx) run_training(hyper_param, model, name_idx=idx, sub_idx=subIdx) autorun_util.close_save_file(orig_stdout, f)
def main(_): hyper_param_list = def_hyper_param() # hyper_param_list = [{'layer': 2, 'feat': [64, 64]}, # {'layer': 3, 'feat': [64, 64, 64]}, # {'layer': 3, 'feat': [32, 16, 16]}] # {'layer': 1, 'feat': [128]}, # {'layer': 2, 'feat': [128, 8]}, # {'layer': 2, 'feat': [128, 16]}, # {'layer': 2, 'feat': [128, 32]}, # # hyper_param_list = [{'layer': 3, 'feat': [4, 4, 512]}, # {'layer': 3, 'feat': [8, 8, 512]}, # {'layer': 3, 'feat': [4, 4, 256]}, # {'layer': 4, 'feat': [4, 4, 4, 512]}, # {'layer': 4, 'feat': [8, 8, 8, 512]}, # {'layer': 4, 'feat': [4, 4, 4, 256]}, # {'layer': 5, 'feat': [4, 4, 4, 4, 512]}, # {'layer': 5, 'feat': [8, 8, 8, 8, 512]}, # {'layer': 5, 'feat': [4, 4, 4, 4, 128]}] model = 7 for hyper_param in hyper_param_list: print("Currently running: ") print("FeatMap: ") print(hyper_param['feat']) print("Model" + str(model)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat']) run_training(hyper_param, model) autorun_util.close_save_file(orig_stdout, f)
def main(_): hyper_param_list = def_hyper_param() for model in range(4, 11): for hyper_param in hyper_param_list: print("Currently running: ") print("FeatMap: ") print(hyper_param['feat']) print("Model" + str(model)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat']) run_training(hyper_param, model) autorun_util.close_save_file(orig_stdout, f)
def main(_): hyper_param_list = def_hyper_param() for model in range(0, 11): for hyper_param in hyper_param_list: print("Currently running: ") print("FeatMap: ") print(hyper_param['feat']) print("Model" + str(model)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat']) run_training(hyper_param, model, isPool=False) # test on no pooling case autorun_util.close_save_file(orig_stdout, f)
def main(_): #hyper_param_list = def_hyper_param() for model in range(0, 1): #for hyper_param in hyper_param_list: hyper_param = {'layer': 3, 'feat': [128, 128, 128]} print("Currently running: ") print("FeatMap: ") print(hyper_param['feat']) print("Model" + str(model)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat']) run_training(hyper_param, model) autorun_util.close_save_file(orig_stdout, f)
def main(_): #hyper_param_list = def_hyper_param() hyper_param_list = [{'layer': 2, 'feat': [layer1_feat, layer2_feat]}] #for model in range(0, 1): # model = autorun_deconv_lasso.DECONV_CVCNN for hyper_param in hyper_param_list: print("Currently running: ") print("FeatMap: ") print(hyper_param['feat']) print("Model" + str(model)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat']) run_training(hyper_param, model) autorun_util.close_save_file(orig_stdout, f)
def main(_): #hyper_param_list = def_hyper_param() hyper_param_list = [{'layer': 3, 'feat': [32, 32, 32]}] #for model in range(0, 1): # model = autorun_deconv_lasso.DECONV_CVCNN for hyper_param in hyper_param_list: print("Currently running: ") print("FeatMap: ") print(hyper_param['feat']) print("Model" + str(model)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat']) run_training(hyper_param, model, name_idx=6, sub_idx=167) # 'sub' and subject 12 autorun_util.close_save_file(orig_stdout, f)
def main(_): #hyper_param_list = def_hyper_param() #hyper_param_list = [{'layer': 3, 'feat': [32, 32, 32]}] hyper_param_list = [{'layer': 2, 'feat': [32, 64]}] #for model in range(0, 1): # model = autorun_deconv_lasso.DECONV_CVCNN for hyper_param in hyper_param_list: print("Currently running: ") print("FeatMap: ") print(hyper_param['feat']) print("Model" + str(model)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat']) run_training(hyper_param, model, name_idx=6, sub_idx=0) # 'sub' and subject 12 autorun_util.close_save_file(orig_stdout, f)
def main(_): # hyper_param_list = def_hyper_param() hyper_param_list = [{'layer': 2, 'feat': [32, 64]}] # {'layer': 1, 'feat': [128]}, # {'layer': 2, 'feat': [128, 8]}, # {'layer': 2, 'feat': [128, 16]}, # {'layer': 2, 'feat': [128, 32]}, # # hyper_param_list = [{'layer': 3, 'feat': [4, 4, 512]}, # {'layer': 3, 'feat': [8, 8, 512]}, # {'layer': 3, 'feat': [4, 4, 256]}, # {'layer': 4, 'feat': [4, 4, 4, 512]}, # {'layer': 4, 'feat': [8, 8, 8, 512]}, # {'layer': 4, 'feat': [4, 4, 4, 256]}, # {'layer': 5, 'feat': [4, 4, 4, 4, 512]}, # {'layer': 5, 'feat': [8, 8, 8, 8, 512]}, # {'layer': 5, 'feat': [4, 4, 4, 4, 128]}] models = [1] for model in models: for hyper_param in hyper_param_list: print("Currently running model: " + str(model)) print("FeatMap: ") print(hyper_param['feat']) # for idx in range(3, len(roi_property.DAT_TYPE_STR)): for idx in range(4, 5): print("Data: " + roi_property.DAT_TYPE_STR[idx]) for subIdx in range(0, 1): print("Subject: " + str(subIdx + 1)) orig_stdout, f = autorun_util.open_save_file( model, hyper_param['feat'], name_idx=idx, sub_idx=subIdx) run_training(hyper_param, model, name_idx=idx, sub_idx=subIdx) autorun_util.close_save_file(orig_stdout, f)
def main(_): # models = [1] # test the DNN-CNN model models = [1] # hyper_param_list = def_hyper_param() # hyper_param_list = [{'layer': 3, 'feat': [128, 32, 32]}] hyper_param_list = [{'layer': 2, 'feat': [32, 64]}] for model in models: for hyper_param in hyper_param_list: print("Currently running model: "+str(model)) print("FeatMap: ") print(hyper_param['feat']) # for idx in range(3, len(roi_property.DAT_TYPE_STR)): for idx in range(4, 5): print("Data: " + roi_property.DAT_TYPE_STR[idx]) # for subIdx in range(107, 108): for subIdx in range(0, 1): print("Subject: " + str(subIdx+1)) orig_stdout, f = autorun_util.open_save_file(model, hyper_param['feat'], name_idx=idx, sub_idx=subIdx) run_training(hyper_param, model, name_idx=idx, sub_idx=subIdx) autorun_util.close_save_file(orig_stdout, f)