utils.py

# Copyright (c) XXXXXXXXXX
# 
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
# 
#     http://www.apache.org/licenses/LICENSE-2.0
# 
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""
Misc functions.

Mostly copy-paste from torchvision references or other public repos like DETR:
https://github.com/facebookresearch/detr/blob/master/util/misc.py
"""
import argparse
import os
import sys
import time
import math
import datetime
import subprocess
import warnings
from ast import literal_eval
from collections import defaultdict, deque, OrderedDict
from functools import wraps

import numpy as np
import torch
from torch import nn
import torch.distributed as dist


def load_pretrained_weights(model, pretrained_weights, checkpoint_key, head=False, head_only=False, strict=False, 
                            prefix=None):
    if not head and head_only:
        raise ValueError("head_only is True but head is False")
    if os.path.isfile(pretrained_weights):
        state_dict = torch.load(pretrained_weights, map_location="cpu")
        if checkpoint_key is not None and checkpoint_key in state_dict:
            state_dict = state_dict[checkpoint_key]
        
        # remove `module.` prefix
        state_dict = {k.replace("module.", ""): v for k, v in state_dict.items()}
        if not head:
            # remove `backbone.` prefix induced by multicrop wrapper    
            state_dict = {k.replace("backbone.", ""): v for k, v in state_dict.items()}
            if prefix is not None:
                new_dict = {}
                for k, v in state_dict.items():
                    if k.startswith(prefix):
                        k_new = k.lstrip(prefix)
                        new_dict[k_new] = v
                state_dict = new_dict
                       
        if head_only:
            state_dict = OrderedDict([(k, v) for k, v in state_dict.items() if not 'backbone' in k])

        msg = model.load_state_dict(state_dict, strict=strict)
        print('Pretrained path {} and loaded with msg: {}'.format(pretrained_weights, msg))
    else:
        print("There is no reference weights available for this model => We use current weights (possibly random).")


def clip_gradients(model, clip):
    norms = []
    for name, p in model.named_parameters():
        if p.grad is not None:
            param_norm = p.grad.data.norm(2)
            norms.append(param_norm.item())
            clip_coef = clip / (param_norm + 1e-6)
            if clip_coef < 1:
                p.grad.data.mul_(clip_coef)
    return norms


def cancel_gradients_last_layer(epoch, model, freeze_last_layer):
    if epoch >= freeze_last_layer:
        return
    for n, p in model.named_parameters():
        if "last_layer" in n:
            p.grad = None


def restart_from_checkpoint(ckp_path, run_variables=None, **kwargs):
    """
    Re-start from checkpoint
    """
    if not os.path.isfile(ckp_path):
        return
    print("Found checkpoint at {}".format(ckp_path))

    # open checkpoint file
    checkpoint = torch.load(ckp_path, map_location="cpu")

    # key is what to look for in the checkpoint file
    # value is the object to load
    # example: {'state_dict': model}
    for key, value in kwargs.items():
        if key in checkpoint and value is not None:
            d = {k.replace("module.", ""): v for k, v in checkpoint[key].items()}
            try:
                msg = value.load_state_dict(d)
                print("=> loaded '{}' from checkpoint '{}' with msg {}".format(key, ckp_path, msg))
            except RuntimeError:
                try:
                    msg = value.load_state_dict(d, strict=False)
                    print("=> loaded '{}' from checkpoint '{}' with msg {}".format(key, ckp_path, msg))
                except ValueError:
                    print("=> failed to load '{}' from checkpoint: '{}'".format(key, ckp_path))
        else:
            print("=> key '{}' not found in checkpoint: '{}'".format(key, ckp_path))

    # re load variable important for the run
    if run_variables is not None:
        for var_name in run_variables:
            if var_name in checkpoint:
                run_variables[var_name] = checkpoint[var_name]


def cosine_scheduler(base_value, final_value, epochs, niter_per_ep, warmup_epochs=0, start_warmup_value=0):
    warmup_schedule = np.array([])
    warmup_iters = warmup_epochs * niter_per_ep
    if warmup_epochs > 0:
        warmup_schedule = np.linspace(start_warmup_value, base_value, warmup_iters)

    iters = np.arange(epochs * niter_per_ep - warmup_iters)
    schedule = final_value + 0.5 * (base_value - final_value) * (1 + np.cos(np.pi * iters / len(iters)))

    schedule = np.concatenate((warmup_schedule, schedule))
    assert len(schedule) == epochs * niter_per_ep
    return schedule


def cosine_scheduler_iter(base_value, final_value, total_iters, warmup_fraction, start_warmup_value=0):
    """
    Same as cosine_scheduler, but with total_iters instead of epochs
    """
    warmup_schedule = np.array([])
    warmup_iters = int(total_iters * warmup_fraction)
    if warmup_iters > 0:
        warmup_schedule = np.linspace(start_warmup_value, base_value, warmup_iters)

    iters = np.arange(total_iters - warmup_iters)
    schedule = final_value + 0.5 * (base_value - final_value) * (1 + np.cos(np.pi * iters / len(iters)))
    schedule = np.concatenate((warmup_schedule, schedule))
    assert len(schedule) == total_iters
    return schedule


def bool_flag(s):
    """
    Parse boolean arguments from the command line.
    """
    FALSY_STRINGS = {"off", "false", "0"}
    TRUTHY_STRINGS = {"on", "true", "1"}
    if s.lower() in FALSY_STRINGS:
        return False
    elif s.lower() in TRUTHY_STRINGS:
        return True
    else:
        raise argparse.ArgumentTypeError("invalid value for a boolean flag")


def fix_random_seeds(seed=31):
    """
    Fix random seeds.
    """
    torch.manual_seed(seed)
    torch.cuda.manual_seed_all(seed)
    np.random.seed(seed)


class SmoothedValue(object):
    """Track a series of values and provide access to smoothed values over a
    window or the global series average.
    """

    def __init__(self, window_size=20, fmt=None):
        if fmt is None:
            fmt = "{median:.6f} ({global_avg:.6f})"
        self.deque = deque(maxlen=window_size)
        self.total = 0.0
        self.count = 0
        self.fmt = fmt

    def is_scalar(self):
        return isinstance(self.total, (float, int))

    def update(self, value, n=1):
        self.deque.append(value)
        self.count += n
        self.total += value * n

    def synchronize_between_processes(self):
        """
        Warning: does not synchronize the deque!
        """
        if not is_dist_avail_and_initialized():
            return
        scalar = not isinstance(self.total, torch.Tensor)
        if scalar:
            total = torch.tensor(self.total, dtype=torch.float64, device="cuda")
        else:
            total = self.total
        count = torch.tensor(self.count, dtype=torch.int64, device="cuda")
        dist.barrier()
        dist.all_reduce(total)
        dist.all_reduce(count)
        self.count = int(count)
        self.total = total.item() if scalar else total

    def deq_to_tensor(self):
        if not isinstance(self.deque[-1], torch.Tensor):
            tensor_list = [torch.tensor(v) for v in self.deque]
        else:
            tensor_list = list(self.deque)
        out = torch.stack(tensor_list)
        if out.ndim == 2:
            out = out.T
        return out

    @property
    def median(self):
        return self.deq_to_tensor().median().item()

    @property
    def avg(self):
        return self.deq_to_tensor().mean().item()

    @property
    def global_avg(self):
        return self.total / self.count

    @property
    def max(self):
        return max(self.deq_to_tensor())

    @property
    def value(self):
        return self.deque[-1]

    def __str__(self):
        # TODO maybe improve this
        if isinstance(self.value, torch.Tensor):
            return '...'
        return self.fmt.format(
            median=self.median,
            avg=self.avg,
            global_avg=self.global_avg,
            max=self.max,
            value=self.value)


def reduce_dict(input_dict, average=True):
    """
    Args:
        input_dict (dict): all the values will be reduced
        average (bool): whether to do average or sum
    Reduce the values in the dictionary from all processes so that all processes
    have the averaged results. Returns a dict with the same fields as
    input_dict, after reduction.
    """
    world_size = get_world_size()
    if world_size < 2:
        return input_dict
    with torch.no_grad():
        names = []
        values = []
        # sort the keys so that they are consistent across processes
        for k in sorted(input_dict.keys()):
            names.append(k)
            values.append(input_dict[k])
        values = torch.stack(values, dim=0)
        dist.all_reduce(values)
        if average:
            values /= world_size
        reduced_dict = {k: v for k, v in zip(names, values)}
    return reduced_dict


class MetricLogger(object):
    def __init__(self, delimiter="\t"):
        self.meters = defaultdict(SmoothedValue)
        self.delimiter = delimiter

    @property
    def scalar_meters(self):
        return {k: v for k, v in self.meters.items() if v.is_scalar()}

    def update(self, **kwargs):
        for k, v in kwargs.items():
            if isinstance(v, torch.Tensor) and v.numel() == 1:
                v = v.item()
                assert isinstance(v, (float, int))
            self.meters[k].update(v)

    def update_raw(self, **kwargs):
        for k, v in kwargs.items():
            self.meters[k].update(v)

    def __getattr__(self, attr):
        if attr in self.meters:
            return self.meters[attr]
        if attr in self.__dict__:
            return self.__dict__[attr]
        raise AttributeError("'{}' object has no attribute '{}'".format(
            type(self).__name__, attr))

    def __str__(self):
        loss_str = []
        for name, meter in self.scalar_meters.items():
            loss_str.append(
                "{}: {}".format(name, str(meter))
            )
        return self.delimiter.join(loss_str)

    def synchronize_between_processes(self):
        for meter in self.meters.values():
            meter.synchronize_between_processes()

    def add_meter(self, name, meter):
        self.meters[name] = meter

    def log_every(self, iterable, print_freq, header=None):
        i = 0
        if not header:
            header = ''
        start_time = time.time()
        end = time.time()
        iter_time = SmoothedValue(fmt='{avg:.6f}')
        data_time = SmoothedValue(fmt='{avg:.6f}')
        space_fmt = ':' + str(len(str(len(iterable)))) + 'd'
        if torch.cuda.is_available():
            log_msg = self.delimiter.join([
                header,
                '[{0' + space_fmt + '}/{1}]',
                'eta: {eta}',
                '{meters}',
                'time: {time}',
                'data: {data}',
                'max mem: {memory:.0f}'
            ])
        else:
            log_msg = self.delimiter.join([
                header,
                '[{0' + space_fmt + '}/{1}]',
                'eta: {eta}',
                '{meters}',
                'time: {time}',
                'data: {data}'
            ])
        MB = 1024.0 * 1024.0
        for obj in iterable:
            data_time.update(time.time() - end)
            yield obj
            iter_time.update(time.time() - end)
            if i % print_freq == 0 or i == len(iterable) - 1:
                eta_seconds = iter_time.global_avg * (len(iterable) - i)
                eta_string = str(datetime.timedelta(seconds=int(eta_seconds)))
                if torch.cuda.is_available():
                    print(log_msg.format(
                        i, len(iterable), eta=eta_string,
                        meters=str(self),
                        time=str(iter_time), data=str(data_time),
                        memory=torch.cuda.max_memory_allocated() / MB))
                else:
                    print(log_msg.format(
                        i, len(iterable), eta=eta_string,
                        meters=str(self),
                        time=str(iter_time), data=str(data_time)))
            i += 1
            end = time.time()
        total_time = time.time() - start_time
        total_time_str = str(datetime.timedelta(seconds=int(total_time)))
        print('{} Total time: {} ({:.6f} s / it)'.format(
            header, total_time_str, total_time / len(iterable)))


def get_sha():
    cwd = os.path.dirname(os.path.abspath(__file__))

    def _run(command):
        return subprocess.check_output(command, cwd=cwd).decode('ascii').strip()
    sha = 'N/A'
    diff = "clean"
    branch = 'N/A'
    try:
        sha = _run(['git', 'rev-parse', 'HEAD'])
        subprocess.check_output(['git', 'diff'], cwd=cwd)
        diff = _run(['git', 'diff-index', 'HEAD'])
        diff = "has uncommited changes" if diff else "clean"
        branch = _run(['git', 'rev-parse', '--abbrev-ref', 'HEAD'])
    except Exception:
        pass
    message = f"sha: {sha}, status: {diff}, branch: {branch}"
    return message


def is_dist_avail_and_initialized():
    if not dist.is_available():
        return False
    if not dist.is_initialized():
        return False
    return True


def get_world_size():
    if not is_dist_avail_and_initialized():
        return 1
    return dist.get_world_size()


def all_reduce(value, *args, **kwargs):
    if is_dist_avail_and_initialized():
        dist.all_reduce(value, *args, **kwargs)


def get_rank():
    if not is_dist_avail_and_initialized():
        return 0
    return dist.get_rank()


def is_main_process():
    return get_rank() == 0


def save_on_master(*args, **kwargs):
    if is_main_process():
        torch.save(*args, **kwargs)


def setup_for_distributed(is_master):
    """
    This function disables printing when not in master process
    """
    import builtins as __builtin__
    builtin_print = __builtin__.print

    def print(*args, **kwargs):
        force = kwargs.pop('force', False)
        if is_master or force:
            builtin_print(*args, **kwargs)

    __builtin__.print = print


def init_distributed_mode(args):
    # launched with torch.distributed.launch
    if 'RANK' in os.environ and 'WORLD_SIZE' in os.environ:
        args.rank = int(os.environ["RANK"])
        args.world_size = int(os.environ['WORLD_SIZE'])
        args.gpu = int(os.environ['LOCAL_RANK'])
    # launched with submitit on a slurm cluster
    elif 'SLURM_PROCID' in os.environ:
        args.rank = int(os.environ['SLURM_PROCID'])
        args.gpu = args.rank % torch.cuda.device_count()
    # launched naively with `python main_dino.py`
    # we manually add MASTER_ADDR and MASTER_PORT to env variables
    elif torch.cuda.is_available():
        print('Will run the code on one GPU.')
        args.rank, args.gpu, args.world_size = 0, 0, 1
        os.environ['MASTER_ADDR'] = '127.0.2.2'
        os.environ['MASTER_PORT'] = '29504'
    else:
        print('Does not support training without GPU.')
        sys.exit(1)

    dist.init_process_group(
        backend="nccl",
        init_method=args.dist_url,
        world_size=args.world_size,
        rank=args.rank,
    )

    torch.cuda.set_device(args.gpu)
    print('| distributed init (rank {}): {}'.format(
        args.rank, args.dist_url), flush=True)
    dist.barrier()
    setup_for_distributed(args.rank == 0)


def accuracy(output, target, topk=(1,)):
    """Computes the accuracy over the k top predictions for the specified values of k"""
    maxk = max(topk)
    batch_size = target.size(0)
    _, pred = output.topk(maxk, 1, True, True)
    pred = pred.t()
    correct = pred.eq(target.reshape(1, -1).expand_as(pred))
    return [correct[:k].reshape(-1).float().sum(0) * 100. / batch_size for k in topk]


def _no_grad_trunc_normal_(tensor, mean, std, a, b):
    # Cut & paste from PyTorch official master until it's in a few official releases - RW
    # Method based on https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf
    def norm_cdf(x):
        # Computes standard normal cumulative distribution function
        return (1. + math.erf(x / math.sqrt(2.))) / 2.

    if (mean < a - 2 * std) or (mean > b + 2 * std):
        warnings.warn("mean is more than 2 std from [a, b] in nn.init.trunc_normal_. "
                      "The distribution of values may be incorrect.",
                      stacklevel=2)

    with torch.no_grad():
        # Values are generated by using a truncated uniform distribution and
        # then using the inverse CDF for the normal distribution.
        # Get upper and lower cdf values
        l = norm_cdf((a - mean) / std)
        u = norm_cdf((b - mean) / std)

        # Uniformly fill tensor with values from [l, u], then translate to
        # [2l-1, 2u-1].
        tensor.uniform_(2 * l - 1, 2 * u - 1)

        # Use inverse cdf transform for normal distribution to get truncated
        # standard normal
        tensor.erfinv_()

        # Transform to proper mean, std
        tensor.mul_(std * math.sqrt(2.))
        tensor.add_(mean)

        # Clamp to ensure it's in the proper range
        tensor.clamp_(min=a, max=b)
        return tensor


def trunc_normal_(tensor, mean=0., std=1., a=-2., b=2.):
    # type: (Tensor, float, float, float, float) -> Tensor
    return _no_grad_trunc_normal_(tensor, mean, std, a, b)


class LARS(torch.optim.Optimizer):
    """
    Almost copy-paste from https://github.com/facebookresearch/barlowtwins/blob/main/main.py
    """
    def __init__(self, params, lr=0, weight_decay=0, momentum=0.9, eta=0.001,
                 weight_decay_filter=None, lars_adaptation_filter=None):
        defaults = dict(lr=lr, weight_decay=weight_decay, momentum=momentum,
                        eta=eta, weight_decay_filter=weight_decay_filter,
                        lars_adaptation_filter=lars_adaptation_filter)
        super().__init__(params, defaults)

    @torch.no_grad()
    def step(self):
        for g in self.param_groups:
            for p in g['params']:
                dp = p.grad

                if dp is None:
                    continue

                if p.ndim != 1:
                    dp = dp.add(p, alpha=g['weight_decay'])

                if p.ndim != 1:
                    param_norm = torch.norm(p)
                    update_norm = torch.norm(dp)
                    one = torch.ones_like(param_norm)
                    q = torch.where(param_norm > 0.,
                                    torch.where(update_norm > 0,
                                                (g['eta'] * param_norm / update_norm), one), one)
                    dp = dp.mul(q)

                param_state = self.state[p]
                if 'mu' not in param_state:
                    param_state['mu'] = torch.zeros_like(p)
                mu = param_state['mu']
                mu.mul_(g['momentum']).add_(dp)

                p.add_(mu, alpha=-g['lr'])


def get_params_groups(model):
    regularized = []
    not_regularized = []
    for name, param in model.named_parameters():
        if not param.requires_grad:
            continue
        # we do not regularize biases nor Norm parameters
        if name.endswith(".bias") or len(param.shape) == 1:
            not_regularized.append(param)
        else:
            regularized.append(param)
    return [{'params': regularized}, {'params': not_regularized, 'weight_decay': 0.}]


def has_batchnorms(model):
    bn_types = (nn.BatchNorm1d, nn.BatchNorm2d, nn.BatchNorm3d, nn.SyncBatchNorm)
    for name, module in model.named_modules():
        if isinstance(module, bn_types):
            return True
    return False


class PCA():
    """
    Class to  compute and apply PCA.
    """
    def __init__(self, dim=256, whit=0.5):
        self.dim = dim
        self.whit = whit
        self.mean = None

    def train_pca(self, cov):
        """
        Takes a covariance matrix (np.ndarray) as input.
        """
        d, v = np.linalg.eigh(cov)
        eps = d.max() * 1e-5
        n_0 = (d < eps).sum()
        if n_0 > 0:
            d[d < eps] = eps

        # total energy
        totenergy = d.sum()

        # sort eigenvectors with eigenvalues order
        idx = np.argsort(d)[::-1][:self.dim]
        d = d[idx]
        v = v[:, idx]

        print("keeping %.2f %% of the energy" % (d.sum() / totenergy * 100.0))

        # for the whitening
        d = np.diag(1. / d**self.whit)

        # principal components
        self.dvt = np.dot(d, v.T)

    def apply(self, x):
        # input is from numpy
        if isinstance(x, np.ndarray):
            if self.mean is not None:
                x -= self.mean
            return np.dot(self.dvt, x.T).T

        # input is from torch and is on GPU
        if x.is_cuda:
            if self.mean is not None:
                x -= torch.cuda.FloatTensor(self.mean)
            return torch.mm(torch.cuda.FloatTensor(self.dvt), x.transpose(0, 1)).transpose(0, 1)

        # input if from torch, on CPU
        if self.mean is not None:
            x -= torch.FloatTensor(self.mean)
        return torch.mm(torch.FloatTensor(self.dvt), x.transpose(0, 1)).transpose(0, 1)


def compute_ap(ranks, nres):
    """
    Computes average precision for given ranked indexes.
    Arguments
    ---------
    ranks : zerro-based ranks of positive images
    nres  : number of positive images
    Returns
    -------
    ap    : average precision
    """

    # number of images ranked by the system
    nimgranks = len(ranks)

    # accumulate trapezoids in PR-plot
    ap = 0

    recall_step = 1. / nres

    for j in np.arange(nimgranks):
        rank = ranks[j]

        if rank == 0:
            precision_0 = 1.
        else:
            precision_0 = float(j) / rank

        precision_1 = float(j + 1) / (rank + 1)

        ap += (precision_0 + precision_1) * recall_step / 2.

    return ap


def compute_map(ranks, gnd, kappas=[]):
    """
    Computes the mAP for a given set of returned results.
         Usage:
           map = compute_map (ranks, gnd)
                 computes mean average precsion (map) only
           map, aps, pr, prs = compute_map (ranks, gnd, kappas)
                 computes mean average precision (map), average precision (aps) for each query
                 computes mean precision at kappas (pr), precision at kappas (prs) for each query
         Notes:
         1) ranks starts from 0, ranks.shape = db_size X #queries
         2) The junk results (e.g., the query itself) should be declared in the gnd stuct array
         3) If there are no positive images for some query, that query is excluded from the evaluation
    """

    map = 0.
    nq = len(gnd) # number of queries
    aps = np.zeros(nq)
    pr = np.zeros(len(kappas))
    prs = np.zeros((nq, len(kappas)))
    nempty = 0

    for i in np.arange(nq):
        qgnd = np.array(gnd[i]['ok'])

        # no positive images, skip from the average
        if qgnd.shape[0] == 0:
            aps[i] = float('nan')
            prs[i, :] = float('nan')
            nempty += 1
            continue

        try:
            qgndj = np.array(gnd[i]['junk'])
        except:
            qgndj = np.empty(0)

        # sorted positions of positive and junk images (0 based)
        pos  = np.arange(ranks.shape[0])[np.in1d(ranks[:,i], qgnd)]
        junk = np.arange(ranks.shape[0])[np.in1d(ranks[:,i], qgndj)]

        k = 0;
        ij = 0;
        if len(junk):
            # decrease positions of positives based on the number of
            # junk images appearing before them
            ip = 0
            while (ip < len(pos)):
                while (ij < len(junk) and pos[ip] > junk[ij]):
                    k += 1
                    ij += 1
                pos[ip] = pos[ip] - k
                ip += 1

        # compute ap
        ap = compute_ap(pos, len(qgnd))
        map = map + ap
        aps[i] = ap

        # compute precision @ k
        pos += 1 # get it to 1-based
        for j in np.arange(len(kappas)):
            kq = min(max(pos), kappas[j]); 
            prs[i, j] = (pos <= kq).sum() / kq
        pr = pr + prs[i, :]

    map = map / (nq - nempty)
    pr = pr / (nq - nempty)

    return map, aps, pr, prs


def multi_scale(samples, model):
    v = None
    for s in [1, 1/2**(1/2), 1/2]:  # we use 3 different scales
        if s == 1:
            inp = samples.clone()
        else:
            inp = nn.functional.interpolate(samples, scale_factor=s, mode='bilinear', align_corners=False)
        feats = model(inp).clone()
        if v is None:
            v = feats
        else:
            v += feats
    v /= 3
    v /= v.norm()
    return v


def kv_pair(s):
    # For extra arparse arguments
    k, v = s.split("=")
    try:
        # v is float/int etc, parse it
        v = literal_eval(v)
    except (ValueError, SyntaxError):
        pass
    return k, v


def _dict_to_kv_pair(d, prefix=''):
    # Transforms an  ordered dict into a key-value pair by using the
    # last value as the value and the rest (concatenated) as the key
    items = list(d.items())
    v = items[-1][1]
    if prefix:
        prefix += '_'
    k = prefix + '_'.join([f'{k}={v}' for k, v in items[:-1]])
    return k, v


def numpy_input(func):
    """
    Decorator to ensure that the input to a function is a numpy array.
    Also disables torch gradient computation, just to be sure.
    """
    def _cast_numpy(x):
        if isinstance(x, torch.Tensor):
            x = x.detach().cpu().numpy()
        return x

    @wraps(func)
    def wrapped_func(*args, **kwargs):
        args = [_cast_numpy(x) for x in args]
        kwargs = {k: _cast_numpy(v) for k, v in kwargs.items()}
        with torch.no_grad():
            return func(*args, **kwargs)
    return wrapped_func


@torch.no_grad()
def apply_batchwise(fn, x, batch_size=1024, device=None, out_device="cpu"):
    """
    Applies a function to a large array in a batchwise manner.
    """
    n = x.shape[0]
    y = []
    if device is None:
        device = out_device
    for i in range(0, n, batch_size):
        xi = x[i:i+batch_size].to(device)
        y.append(fn(xi).to(out_device))
    return torch.cat(y, dim=0)


def freeze_transformer_blocks(model, n_blocks=-1):
    """
    Freeze the first n_blocks of the transformer architecture.
    """
    if n_blocks <= 0:
        return model
    c = 0
    p = 0
    for (name, param) in (model.named_parameters()):
        if "block"  in name:
            block_id= int(name.split(".")[1])
            c+=1
        else:
            block_id = 0
        if block_id < n_blocks:
            param.requires_grad = False
            p+=1
        elif block_id >= n_blocks:
            
            param.requires_grad = True
    print(f"Freezing {p} parameters in the first {n_blocks} blocks out of {c} parameters : percentage {p/c}")    
    return model


def freeze_custom_model(model, arch, percent=50):
    """
    Freezes the first <percent>% of the model.
    Only works for archs in 'custom_model_frozen_submodules'
    """
    assert arch in custom_model_frozen_submodules, f"Architecture {arch} not supported for freezing."
    freeze_mods = custom_model_frozen_submodules[arch](model, percent)
    return _freeze_module_list(model, freeze_mods)


def _freeze_module_list(model, module_list):
    num_params_before = sum(p.numel() for p in model.parameters() if p.requires_grad)
    for m in module_list:
        for p in m.parameters():
            p.requires_grad = False
    num_params_after = sum(p.numel() for p in model.parameters() if p.requires_grad)
    print(f"Freezing {(num_params_before - num_params_after) / num_params_before * 100:.2f}% of the model "
          f"({num_params_before - num_params_after} out of {num_params_before} parameters).")
    return model


def partial_freeze(model, percent, dummy_shape=(1, 3, 224, 224)):
    """
    Freezes the first <percent>% of the model.
    To be precise, it freezes the first layers of the model until the desired percentage of parameters would be exceeded.
    """
    freeze_hook = _FreezeHook(model, percent)
    dummy_inp = torch.zeros(dummy_shape, device=next(model.parameters()).device)
    with torch.no_grad():
        model(dummy_inp)
    freeze_hook.remove()
    return _freeze_module_list(model, freeze_hook.modules_to_freeze)


class _FreezeHook:
    def __init__(self, model, percent):
        self.modules_to_freeze = []
        self.num_frozen_params = 0
        self.params_before = sum(p.numel() for p in model.parameters() if p.requires_grad)
        self.num_params_to_freeze = int(self.params_before * percent / 100)
        base_modules = [m for m in model.modules() if len(list(m.children())) == 0]
        self.handles = [m.register_forward_hook(self) for m in base_modules]
        self.seen_modules = set()

    def __call__(self, module, inp, outp):
        if module in self.seen_modules:
            return
        nump = sum(p.numel() for p in module.parameters() if p.requires_grad)
        if self.num_frozen_params + nump <= self.num_params_to_freeze:
            self.modules_to_freeze.append(module)
            self.num_frozen_params += nump
        self.seen_modules.add(module)

    def remove(self):
        for h in self.handles:
            h.remove()


def _convnext_base_parts(model, percent):
    if percent == 25:
        return [model.stem,
                model.stages[:2],
                model.stages[2].downsample,
                model.stages[2].blocks[:9]]
    elif percent == 50:
        return [model.stem,
                model.stages[:2],
                model.stages[2].downsample,
                model.stages[2].blocks[:19]]
    elif percent == 75:
        return [model.stem,
                model.stages[:3],
                model.stages[3].downsample,
                model.stages[3].blocks[0].norm,
                model.stages[3].blocks[0].mlp.fc1]
    raise ValueError(f"Percent {percent} not supported for ConvNext.")


# Manually checked, which submodules to freeze for each model
custom_model_frozen_submodules = {
    'timm_convnext_base_in22k': _convnext_base_parts
}