def init_params(self): for param in self.parameters(): if param.requires_grad: if cfg.use_truncated_normal: truncated_normal_(param, std=0.1) else: torch.nn.init.normal_(param, std=0.1)
def init_params(self): for param in self.parameters(): if param.requires_grad and len(param.shape) > 0: stddev = 1 / math.sqrt(param.shape[0]) if cfg.use_truncated_normal: truncated_normal_(param, std=stddev) else: torch.nn.init.normal_(param, std=stddev)
def init_params(self): for param in self.parameters(): if param.requires_grad and len(param.shape) > 0: stddev = 1 / math.sqrt(param.shape[0]) if cfg.dis_init == 'uniform': torch.nn.init.uniform_(param, a=-0.05, b=0.05) elif cfg.dis_init == 'normal': torch.nn.init.normal_(param, std=stddev) elif cfg.dis_init == 'truncated_normal': truncated_normal_(param, std=stddev)