optim_factory.py

import torch
from torch import optim as optim

# Optional timm optimizers: keep imports version-tolerant.
try:
    from timm.optim.adafactor import Adafactor
except Exception:
    Adafactor = None

try:
    from timm.optim.adahessian import Adahessian
except Exception:
    Adahessian = None

try:
    from timm.optim.adamp import AdamP
except Exception:
    AdamP = None

try:
    from timm.optim.lookahead import Lookahead
except Exception:
    Lookahead = None

try:
    from timm.optim.nadam import Nadam
except Exception:
    try:
        # timm>=1.x may expose `NAdam` instead of `Nadam`
        from timm.optim.nadam import NAdam as Nadam
    except Exception:
        Nadam = None

try:
    from timm.optim.radam import RAdam
except Exception:
    RAdam = None

try:
    from timm.optim.rmsprop_tf import RMSpropTF
except Exception:
    RMSpropTF = None

try:
    from timm.optim.sgdp import SGDP
except Exception:
    SGDP = None

import json

try:
    from apex.optimizers import FusedNovoGrad, FusedAdam, FusedLAMB, FusedSGD
    has_apex = True
except ImportError:
    has_apex = False


def get_num_layer_for_vit(var_name, num_max_layer):
    if var_name in ("cls_token", "mask_token", "pos_embed"):
        return 0
    elif var_name.startswith("patch_embed"):
        return 0
    elif var_name.startswith("rel_pos_bias"):
        return num_max_layer - 1
    elif var_name.startswith("blocks"):
        layer_id = int(var_name.split('.')[1])
        return layer_id + 1
    else:
        return num_max_layer - 1


class LayerDecayValueAssigner(object):
    def __init__(self, values):
        self.values = values

    def get_scale(self, layer_id):
        return self.values[layer_id]

    def get_layer_id(self, var_name):
        return get_num_layer_for_vit(var_name, len(self.values))


def get_parameter_groups(model, weight_decay=1e-5, skip_list=(), get_num_layer=None, get_layer_scale=None):
    parameter_group_names = {}
    parameter_group_vars = {}

    for name, param in model.named_parameters():
        if not param.requires_grad:
            continue  # frozen weights
        if len(param.shape) == 1 or name.endswith(".bias") or name in skip_list:
            group_name = "no_decay"
            this_weight_decay = 0.
        else:
            group_name = "decay"
            this_weight_decay = weight_decay
        if get_num_layer is not None:
            layer_id = get_num_layer(name)
            group_name = "layer_%d_%s" % (layer_id, group_name)
        else:
            layer_id = None

        if group_name not in parameter_group_names:
            if get_layer_scale is not None:
                scale = get_layer_scale(layer_id)
            else:
                scale = 1.

            parameter_group_names[group_name] = {
                "weight_decay": this_weight_decay,
                "params": [],
                "lr_scale": scale
            }
            parameter_group_vars[group_name] = {
                "weight_decay": this_weight_decay,
                "params": [],
                "names": [],
                "lr_scale": scale
            }

        parameter_group_vars[group_name]["params"].append(param)
        parameter_group_vars[group_name]["names"].append(name)
        parameter_group_names[group_name]["params"].append(name)
    print("Param groups = %s" % json.dumps(parameter_group_names, indent=2))
    return list(parameter_group_vars.values())


def create_optimizer(args, model, get_num_layer=None, get_layer_scale=None, filter_bias_and_bn=True, skip_list=None):
    opt_lower = args.opt.lower()
    weight_decay = args.weight_decay
    if weight_decay and filter_bias_and_bn:
        skip = {}
        if skip_list is not None:
            skip = skip_list
        elif hasattr(model, 'no_weight_decay'):
            skip = model.no_weight_decay()
        parameters = get_parameter_groups(model, weight_decay, skip, get_num_layer, get_layer_scale)
        weight_decay = 0.
    else:
        parameters = model.parameters()

    if 'fused' in opt_lower:
        assert has_apex and torch.cuda.is_available(), 'APEX and CUDA required for fused optimizers'

    opt_args = dict(lr=args.lr, weight_decay=weight_decay)
    if hasattr(args, 'opt_eps') and args.opt_eps is not None:
        opt_args['eps'] = args.opt_eps
    if hasattr(args, 'opt_betas') and args.opt_betas is not None:
        opt_args['betas'] = args.opt_betas

    print("optimizer settings:", opt_args)

    opt_split = opt_lower.split('_')
    opt_lower = opt_split[-1]
    if opt_lower == 'sgd' or opt_lower == 'nesterov':
        opt_args.pop('eps', None)
        optimizer = optim.SGD(parameters, momentum=args.momentum, nesterov=True, **opt_args)
    elif opt_lower == 'momentum':
        opt_args.pop('eps', None)
        optimizer = optim.SGD(parameters, momentum=args.momentum, nesterov=False, **opt_args)
    elif opt_lower == 'adam':
        optimizer = optim.Adam(parameters, **opt_args)
    elif opt_lower == 'adamw':
        optimizer = optim.AdamW(parameters, **opt_args)
    # elif opt_lower == 'nadam':
    #     optimizer = Nadam(parameters, **opt_args)
    # elif opt_lower == 'radam':
    #     optimizer = RAdam(parameters, **opt_args)
    # elif opt_lower == 'adamp':
    #     optimizer = AdamP(parameters, wd_ratio=0.01, nesterov=True, **opt_args)
    # elif opt_lower == 'sgdp':
    #     optimizer = SGDP(parameters, momentum=args.momentum, nesterov=True, **opt_args)
    # elif opt_lower == 'adadelta':
    #     optimizer = optim.Adadelta(parameters, **opt_args)
    # elif opt_lower == 'adafactor':
    #     if not args.lr:
    #         opt_args['lr'] = None
    #     optimizer = Adafactor(parameters, **opt_args)
    # elif opt_lower == 'adahessian':
    #     optimizer = Adahessian(parameters, **opt_args)
    # elif opt_lower == 'rmsprop':
    #     optimizer = optim.RMSprop(parameters, alpha=0.9, momentum=args.momentum, **opt_args)
    # elif opt_lower == 'rmsproptf':
    #     optimizer = RMSpropTF(parameters, alpha=0.9, momentum=args.momentum, **opt_args)
    # elif opt_lower == 'novograd':
    #     optimizer = NovoGrad(parameters, **opt_args)
    # elif opt_lower == 'nvnovograd':
    #     optimizer = NvNovoGrad(parameters, **opt_args)
    # elif opt_lower == 'fusedsgd':
    #     opt_args.pop('eps', None)
    #     optimizer = FusedSGD(parameters, momentum=args.momentum, nesterov=True, **opt_args)
    # elif opt_lower == 'fusedmomentum':
    #     opt_args.pop('eps', None)
    #     optimizer = FusedSGD(parameters, momentum=args.momentum, nesterov=False, **opt_args)
    # elif opt_lower == 'fusedadam':
    #     optimizer = FusedAdam(parameters, adam_w_mode=False, **opt_args)
    # elif opt_lower == 'fusedadamw':
    #     optimizer = FusedAdam(parameters, adam_w_mode=True, **opt_args)
    # elif opt_lower == 'fusedlamb':
    #     optimizer = FusedLAMB(parameters, **opt_args)
    # elif opt_lower == 'fusednovograd':
    #     opt_args.setdefault('betas', (0.95, 0.98))
    #     optimizer = FusedNovoGrad(parameters, **opt_args)
    else:
        assert False and "Invalid optimizer"
        raise ValueError

    if len(opt_split) > 1:
        if opt_split[0] == 'lookahead':
            if Lookahead is None:
                raise RuntimeError("Lookahead optimizer is not available in the installed timm version.")
            optimizer = Lookahead(optimizer)

    return optimizer