U ʹj@@sddlZddlZddlmZddlZddlZddlmZddlm Z ddl m Z dddd d d d gZ dd l m Z ddd Zdd Zddd Zdd ZGdddeZedddZGddde ZdS)N)deepcopy)Module) LRScheduler)&_get_foreach_kernels_supported_devices AveragedModel update_bnSWALRget_ema_multi_avg_fnget_swa_multi_avg_fnget_ema_avg_fnget_swa_avg_fn)"_group_tensors_by_device_and_dtype+?cstfdd}|S)Ncsbt|dst|dr0t||dn.t||D]"\}}|||dq:dSNr)torchis_floating_point is_complex_foreach_lerp_zipcopy_)Zema_param_listcurrent_param_list_Zp_emap_modeldecayT/var/www/html/bwave/python/venv/lib/python3.8/site-packages/torch/optim/swa_utils.py ema_updatesz(get_ema_multi_avg_fn..ema_updaterno_gradrrrrrr scCstdd}|S)NcSs`t|dst|dr4t||d|dn(t||}t|||dgt|dSr)rrrrZ _foreach_subZ_foreach_addcdiv_len)Zaveraged_param_listr num_averagedZdiffsrrr swa_update&s z(get_swa_multi_avg_fn..swa_updaterr$rrrr %s cstfdd}|S)Ncs|d|SNrr)Z ema_param current_paramr#rrrr3sz"get_ema_avg_fn..ema_updaterr!rrrr 2scCstdd}|S)NcSs||||dSr&r)Zaveraged_paramr'r#rrrr$;sz"get_swa_avg_fn..swa_updaterr%rrrr :s cs2eZdZdZd fdd ZddZdd ZZS) raImplements averaged model for Stochastic Weight Averaging (SWA) and Exponential Moving Average (EMA). Stochastic Weight Averaging was proposed in `Averaging Weights Leads to Wider Optima and Better Generalization`_ by Pavel Izmailov, Dmitrii Podoprikhin, Timur Garipov, Dmitry Vetrov and Andrew Gordon Wilson (UAI 2018). Exponential Moving Average is a variation of `Polyak averaging`_, but using exponential weights instead of equal weights across iterations. AveragedModel class creates a copy of the provided module :attr:`model` on the device :attr:`device` and allows to compute running averages of the parameters of the :attr:`model`. Args: model (torch.nn.Module): model to use with SWA/EMA device (torch.device, optional): if provided, the averaged model will be stored on the :attr:`device` avg_fn (function, optional): the averaging function used to update parameters; the function must take in the current value of the :class:`AveragedModel` parameter, the current value of :attr:`model` parameter, and the number of models already averaged; if None, an equally weighted average is used (default: None) multi_avg_fn (function, optional): the averaging function used to update parameters inplace; the function must take in the current values of the :class:`AveragedModel` parameters as a list, the current values of :attr:`model` parameters as a list, and the number of models already averaged; if None, an equally weighted average is used (default: None) use_buffers (bool): if ``True``, it will compute running averages for both the parameters and the buffers of the model. (default: ``False``) Example: >>> # xdoctest: +SKIP("undefined variables") >>> loader, optimizer, model, loss_fn = ... >>> swa_model = torch.optim.swa_utils.AveragedModel(model) >>> scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(optimizer, >>> T_max=300) >>> swa_start = 160 >>> swa_scheduler = SWALR(optimizer, swa_lr=0.05) >>> for i in range(300): >>> for input, target in loader: >>> optimizer.zero_grad() >>> loss_fn(model(input), target).backward() >>> optimizer.step() >>> if i > swa_start: >>> swa_model.update_parameters(model) >>> swa_scheduler.step() >>> else: >>> scheduler.step() >>> >>> # Update bn statistics for the swa_model at the end >>> torch.optim.swa_utils.update_bn(loader, swa_model) You can also use custom averaging functions with the `avg_fn` or `multi_avg_fn` parameters. If no averaging function is provided, the default is to compute equally-weighted average of the weights (SWA). Example: >>> # xdoctest: +SKIP("undefined variables") >>> # Compute exponential moving averages of the weights and buffers >>> ema_model = torch.optim.swa_utils.AveragedModel(model, >>> torch.optim.swa_utils.get_ema_multi_avg_fn(0.9), use_buffers=True) .. note:: When using SWA/EMA with models containing Batch Normalization you may need to update the activation statistics for Batch Normalization. This can be done either by using the :meth:`torch.optim.swa_utils.update_bn` or by setting :attr:`use_buffers` to `True`. The first approach updates the statistics in a post-training step by passing data through the model. The second does it during the parameter update phase by averaging all buffers. Empirical evidence has shown that updating the statistics in normalization layers increases accuracy, but you may wish to empirically test which approach yields the best results in your problem. .. note:: :attr:`avg_fn` and `multi_avg_fn` are not saved in the :meth:`state_dict` of the model. .. note:: When :meth:`update_parameters` is called for the first time (i.e. :attr:`n_averaged` is `0`) the parameters of `model` are copied to the parameters of :class:`AveragedModel`. For every subsequent call of :meth:`update_parameters` the function `avg_fn` is used to update the parameters. .. _Averaging Weights Leads to Wider Optima and Better Generalization: https://arxiv.org/abs/1803.05407 .. _There Are Many Consistent Explanations of Unlabeled Data: Why You Should Average: https://arxiv.org/abs/1806.05594 .. _SWALP: Stochastic Weight Averaging in Low-Precision Training: https://arxiv.org/abs/1904.11943 .. _Stochastic Weight Averaging in Parallel: Large-Batch Training That Generalizes Well: https://arxiv.org/abs/2001.02312 .. _Polyak averaging: https://paperswithcode.com/method/polyak-averaging NFcsrt|dks"|dks"tdt||_|dk rB|j||_|dtjdtj |d||_ ||_ ||_ dS)Nz6Only one of avg_fn and multi_avg_fn should be provided n_averagedr)Zdtypedevice) super__init__AssertionErrorrmoduletoZregister_bufferrZtensorlongavg_fn multi_avg_fn use_buffers)selfmodelr)r0r1r2 __class__rrr+s  zAveragedModel.__init__cOs |j||SN)r-)r3argskwargsrrrforwardszAveragedModel.forwardcCs |jrt|j|jn|}|jr@t||n|}g}g}t||D]H\}}||j }| || ||j dkrZ| |qZ|j dkr|j dk s|jdkrvt||g} | D]\\} } \\} } } |j r| | | |j | q| jtkr8t}|| | |j | qt}|j | }t| | D]\}}| ||||qTqn@t||D]4\}}|j |j }| ||||q|jst|j|D]$\}}| ||j q|j d7_ dSr)r2 itertoolschainr- parametersbuffersrdetachr.r)appendr(rr1r0r itemstyperr r )r3r4Z self_paramZ model_paramZself_param_detachedZmodel_param_detachedZ p_averagedrZp_model_Zgrouped_tensorsr)rZ self_paramsZ model_paramsr1r0r(Zb_swaZb_modelrrrupdate_parameterssH      zAveragedModel.update_parameters)NNNF)__name__ __module__ __qualname____doc__r+r:rC __classcell__rrr5rrBsb cCsi}|D](}t|tjjjjr ||j||<q |s>dS|j}| | D] }d|_qT|D]4}t|t t fr~|d}|dk r| |}||qd| D]}|||_q| |dS)aUpdates BatchNorm running_mean, running_var buffers in the model. It performs one pass over data in `loader` to estimate the activation statistics for BatchNorm layers in the model. Args: loader (torch.utils.data.DataLoader): dataset loader to compute the activation statistics on. Each data batch should be either a tensor, or a list/tuple whose first element is a tensor containing data. model (torch.nn.Module): model for which we seek to update BatchNorm statistics. device (torch.device, optional): If set, data will be transferred to :attr:`device` before being passed into :attr:`model`. Example: >>> # xdoctest: +SKIP("Undefined variables") >>> loader, model = ... >>> torch.optim.swa_utils.update_bn(loader, model) .. note:: The `update_bn` utility assumes that each data batch in :attr:`loader` is either a tensor or a list or tuple of tensors; in the latter case it is assumed that :meth:`model.forward()` should be called on the first element of the list or tuple corresponding to the data batch. Nr)modules isinstancernnZ batchnormZ _BatchNormZreset_running_statsZmomentumZtrainingtrainkeyslisttupler.)loaderr4r)Zmomentar-Z was_traininginputZ bn_modulerrrrs(       csZeZdZdZdfdd ZeddZed d Zed d Zed dZ ddZ Z S)raDAnneals the learning rate in each parameter group to a fixed value. This learning rate scheduler is meant to be used with Stochastic Weight Averaging (SWA) method (see `torch.optim.swa_utils.AveragedModel`). Args: optimizer (torch.optim.Optimizer): wrapped optimizer swa_lrs (float or list): the learning rate value for all param groups together or separately for each group. annealing_epochs (int): number of epochs in the annealing phase (default: 10) annealing_strategy (str): "cos" or "linear"; specifies the annealing strategy: "cos" for cosine annealing, "linear" for linear annealing (default: "cos") last_epoch (int): the index of the last epoch (default: -1) The :class:`SWALR` scheduler can be used together with other schedulers to switch to a constant learning rate late in the training as in the example below. Example: >>> # xdoctest: +SKIP("Undefined variables") >>> loader, optimizer, model = ... >>> lr_lambda = lambda epoch: 0.9 >>> scheduler = torch.optim.lr_scheduler.MultiplicativeLR(optimizer, >>> lr_lambda=lr_lambda) >>> swa_scheduler = torch.optim.swa_utils.SWALR(optimizer, >>> anneal_strategy="linear", anneal_epochs=20, swa_lr=0.05) >>> swa_start = 160 >>> for i in range(300): >>> for input, target in loader: >>> optimizer.zero_grad() >>> loss_fn(model(input), target).backward() >>> optimizer.step() >>> if i > swa_start: >>> swa_scheduler.step() >>> else: >>> scheduler.step() .. _Averaging Weights Leads to Wider Optima and Better Generalization: https://arxiv.org/abs/1803.05407 coscs|||}t||jD]\}}||d<q|dkrBtd|n"|dkrT|j|_n|dkrd|j|_t|trv|dkrtd|||_ t ||dS)Nswa_lr)rSlinearz>anneal_strategy must by one of 'cos' or 'linear', instead got rSrVrz3anneal_epochs must be equal or greater than 0, got ) _format_paramr param_groups ValueError_cosine_anneal anneal_func_linear_annealrJint anneal_epochsr*r+)r3 optimizerrUr^Zanneal_strategyZ last_epochswa_lrsgroupr5rrr+Bs   zSWALR.__init__cCsVt|ttfrBt|t|jkr>tdt|dt|j|S|gt|jSdS)NzGswa_lr must have the same length as optimizer.param_groups: swa_lr has z, optimizer.param_groups has )rJrNrOr"rXrY)r_r`rrrrWRs zSWALR._format_paramcCs|Sr7rtrrrr\]szSWALR._linear_annealcCsdttj|dS)Nr)mathrSpirbrrrrZaszSWALR._cosine_annealcCs |dkr |S|||d|Sr&r)lrrUalpharrr_get_initial_lreszSWALR._get_initial_lrcsjstdtjd}jdkr0td|}tdtd|dtdj}|fddj j D}tdtd|tdj}|fddt j j |DS)NzTTo get the last learning rate computed by the scheduler, please use `get_last_lr()`.rrcs"g|]}|d|dqS)rgrU)ri).0ra) prev_alphar3rr tsz SWALR.get_lr..cs(g|] \}}|d|dqS)rUrr)rjrarg)rhrrrlxs) Z_get_lr_called_within_stepwarningswarn UserWarningZ _step_countr^maxminr[r_rXr)r3stepZprev_tZprev_lrsrcr)rhrkr3rget_lrks"         z SWALR.get_lr)rRrSrT) rDrErFrGr+ staticmethodrWr\rZrirsrHrrr5rrs*    )r)r)N)r;recopyrrmrZtorch.nnrZtorch.optim.lr_schedulerrZtorch.utils._foreach_utilsr__all__r r r r r rr rrrrrrs2         5