linna.nnutils

Module Contents

Classes

Params	Class that loads hyperparameters from a json file.
RunningAverage	A simple class that maintains the running average of a quantity

Functions

set_logger(log_path)	Set the logger to log info in terminal and file log_path.
save_dict_to_json(d, json_path)	Saves dict of floats in json file
save_checkpoint(state, is_best, checkpoint)	Saves model and training parameters at checkpoint + 'last.pth.tar'. If is_best==True, also saves
load_checkpoint(checkpoint, model, optimizer=None, device=None, ismpi=False)	Loads model parameters (state_dict) from file_path. If optimizer is provided, loads state_dict of
gen_plot(plotarr, shape)	Create a pyplot plot and save to buffer.

class linna.nnutils.Params(json_path)

Class that loads hyperparameters from a json file. Example: ` params = Params(json_path) print(params.learning_rate) params.learning_rate = 0.5  # change the value of learning_rate in params `

save(self, json_path)

update(self, json_path)

Loads parameters from json file

property dict(self)

Gives dict-like access to Params instance by `params.dict[‘learning_rate’]

class linna.nnutils.RunningAverage

A simple class that maintains the running average of a quantity

Example: ` loss_avg = RunningAverage() loss_avg.update(2) loss_avg.update(4) loss_avg() = 3 `

update(self, val)

__call__(self)

linna.nnutils.set_logger(log_path)

Set the logger to log info in terminal and file log_path. In general, it is useful to have a logger so that every output to the terminal is saved in a permanent file. Here we save it to model_dir/train.log. Example: ` logging.info("Starting training...") ` :param log_path: (string) where to log

linna.nnutils.save_dict_to_json(d, json_path)

Saves dict of floats in json file :param d: (dict) of float-castable values (np.float, int, float, etc.) :param json_path: (string) path to json file

linna.nnutils.save_checkpoint(state, is_best, checkpoint)

Saves model and training parameters at checkpoint + ‘last.pth.tar’. If is_best==True, also saves checkpoint + ‘best.pth.tar’ :param state: (dict) contains model’s state_dict, may contain other keys such as epoch, optimizer state_dict :param is_best: (bool) True if it is the best model seen till now :param checkpoint: (string) folder where parameters are to be saved

linna.nnutils.load_checkpoint(checkpoint, model, optimizer=None, device=None, ismpi=False)

Loads model parameters (state_dict) from file_path. If optimizer is provided, loads state_dict of optimizer assuming it is present in checkpoint. :param checkpoint: (string) filename which needs to be loaded :param model: (torch.nn.Module) model for which the parameters are loaded :param optimizer: (torch.optim) optional: resume optimizer from checkpoint

linna.nnutils.gen_plot(plotarr, shape)

Create a pyplot plot and save to buffer.