集群训练bug记录

-- Process 0 terminated with the following error:
Traceback (most recent call last):
  File "/opt/conda/lib/python3.7/site-packages/torch/multiprocessing/spawn.py", line 20, in _wrap
    fn(i, *args)
  File "/workspace/geniii-trainingcode-lane/lanenet/distribute_prune_erfnet_cluster.py", line 628, in _distributed_worker
    main_func()
  File "/workspace/geniii-trainingcode-lane/lanenet/distribute_prune_erfnet_cluster.py", line 234, in main
    prune_statistics = pruner.prune_model(model, 0, input_shape)  # iteration=0
  File "/workspace/geniii-trainingcode-lane/lanenet/ambasparsificationtool/pytorch/sparsification/ptpruner/Pytorch_pruner.py", line 245, in prune_model
    self.sparse_analysis(model, input_shape)
  File "/workspace/geniii-trainingcode-lane/lanenet/ambasparsificationtool/pytorch/sparsification/ptpruner/Pytorch_pruner.py", line 792, in sparse_analysis
    curr_sparse = self._parameter_hist(weight_info_dict, model, curr_energy, self.sparse_table[len(desire_index_table)], tmp_matrix, b_achieve_net_max_sparse, self.prun_algo, self.prun_algo_tuning, self.no_prun_keyword, self.no_prun_layer, self.bin_core_sparse_ratio_decision, layer_info, module_name_map)
  File "/workspace/geniii-trainingcode-lane/lanenet/ambasparsificationtool/pytorch/sparsification/ptpruner/Pytorch_pruner.py", line 725, in _parameter_hist
    name, ctrl_bits, is_dw, prun_algo, prun_algo_tuning, bin_core_sparse_ratio_decision, weight_info_dict)
  File "/workspace/geniii-trainingcode-lane/lanenet/ambasparsificationtool/pytorch/sparsification/ptpruner/Pytorch_pruner.py", line 634, in _sparse_ratio_decision
    energy_th = float(subprocess.Popen('{} {}'.format(bin_core_sparse_ratio_decision, argc_list), shell=True, stdout=subprocess.PIPE).communicate()[0])
ValueError: could not convert string to float: