llava/train/webdataset_utils.py

import ast
import json
import logging
import math
import os
import random
import sys
from dataclasses import dataclass
from multiprocessing import Value

import braceexpand
import numpy as np
import pandas as pd
import torch
import webdataset as wds
from PIL import Image
from torch.utils.data import  DataLoader,  IterableDataset, get_worker_info
from webdataset.filters import _shuffle
from webdataset.tariterators import base_plus_ext, url_opener, tar_file_expander, valid_sample
# from data_utils import get_normalized_weights_and_num_samples
from typing import List, Tuple


def get_normalized_weights_and_num_samples(
    weights: List[float], num_samples: int
) -> Tuple[List[float], List[int]]:
    # Normalize weights
    weight_sum = sum(weights)
    assert weight_sum > 0.0
    weights = [weight / weight_sum for weight in weights]
    # Add 0.5% (the 1.005 factor) so in case the blending dataset does
    # not uniformly distribute the number of samples, we still have
    # samples left to feed to the network.
    weighted_num_samples = []
    for weight in weights:
        weighted_num_samples.append(int(math.ceil(num_samples * weight * 1.005)))
    return weights, weighted_num_samples

class SharedEpoch:
    def __init__(self, epoch: int = 0):
        self.shared_epoch = Value('i', epoch)

    def set_value(self, epoch):
        self.shared_epoch.value = epoch

    def get_value(self):
        return self.shared_epoch.value


@dataclass
class DataInfo:
    dataloader: DataLoader
    shared_epoch: SharedEpoch = None

    def set_epoch(self, epoch):
        if self.shared_epoch is not None:
            self.shared_epoch.set_value(epoch)

def expand_urls(urls, weights=None):
    if weights is None:
        expanded_urls = wds.shardlists.expand_urls(urls)
        expanded_urls=[url for url in expanded_urls if os.path.exists(url)]
        ### go save existed url
        return expanded_urls, None
    if isinstance(urls, str):
        urllist = urls.split("::")
        weights = weights.split('::')
        assert len(weights) == len(urllist),\
            f"Expected the number of data components ({len(urllist)}) and weights({len(weights)}) to match."
        weights = [float(weight) for weight in weights]
        all_urls, all_weights = [], []
        for url, weight in zip(urllist, weights):
            expanded_url = list(braceexpand.braceexpand(url))
            expanded_urls=[url for url in expanded_urls if os.path.exists(url)] 
            ### go save existed url
            expanded_weights = [weight for _ in expanded_url]
            all_urls.extend(expanded_url)
            all_weights.extend(expanded_weights)
        return all_urls, all_weights
    else:
        all_urls = list(urls)
        return all_urls, weights


def get_dataset_size(shards):
    shards_list = (shards)
    dir_path = os.path.dirname(shards_list[0])
    sizes_filename = os.path.join(dir_path, 'sizes.json')
    len_filename = os.path.join(dir_path, '__len__')
    if os.path.exists(sizes_filename):
        sizes = json.load(open(sizes_filename, 'r'))
        total_size = sum([int(sizes[os.path.basename(shard)]) for shard in shards_list])
    elif os.path.exists(len_filename):
        # FIXME this used to be eval(open(...)) but that seemed rather unsafe
        total_size = ast.literal_eval(open(len_filename, 'r').read())
    else:
        total_size = None  # num samples undefined
        # some common dataset sizes (at time of authors last download)
        # CC3M (train): 2905954
        # CC12M: 10968539
        # LAION-400M: 407332084
        # LAION-2B (english): 2170337258
    num_shards = len(shards_list)
    return total_size, num_shards

def count_samples(dataloader):
    os.environ["WDS_EPOCH"] = "0"
    n_elements, n_batches = 0, 0
    for images, texts in dataloader:
        n_batches += 1
        n_elements += len(images)
        assert len(images) == len(texts)
    return n_elements, n_batches

def filter_no_caption_or_no_image(sample):
    has_caption = ('txt' in sample)
    has_image = ('png' in sample or 'jpg' in sample or 'jpeg' in sample or 'webp' in sample)
    return has_caption and has_image


def log_and_continue(exn):
    """Call in an exception handler to ignore any exception, issue a warning, and continue."""
    logging.warning(f'Handling webdataset error ({repr(exn)}). Ignoring.')
    return True

def group_by_keys_nothrow(data, keys=base_plus_ext, lcase=True, suffixes=None, handler=None):
    """Return function over iterator that groups key, value pairs into samples.

    :param keys: function that splits the key into key and extension (base_plus_ext)
    :param lcase: convert suffixes to lower case (Default value = True)
    """
    current_sample = None
    for filesample in data:
        assert isinstance(filesample, dict)
        fname, value = filesample["fname"], filesample["data"]
        prefix, suffix = keys(fname)
        if prefix is None:
            continue
        if lcase:
            suffix = suffix.lower()
        # FIXME webdataset version throws if suffix in current_sample, but we have a potential for
        #  this happening in the current LAION400m dataset if a tar ends with same prefix as the next
        #  begins, rare, but can happen since prefix aren't unique across tar files in that dataset
        if current_sample is None or prefix != current_sample["__key__"] or suffix in current_sample:
            if valid_sample(current_sample):
                yield current_sample
            current_sample = dict(__key__=prefix, __url__=filesample["__url__"])
        if suffixes is None or suffix in suffixes:
            current_sample[suffix] = value
    if valid_sample(current_sample):
        yield current_sample


def tarfile_to_samples_nothrow(src, handler=log_and_continue):
    # NOTE this is a re-impl of the webdataset impl with group_by_keys that doesn't throw
    streams = url_opener(src, handler=handler)
    files = tar_file_expander(streams, handler=handler)
    samples = group_by_keys_nothrow(files, handler=handler)
    return samples


def pytorch_worker_seed(increment=0):
    """get dataloader worker seed from pytorch"""
    worker_info = get_worker_info()
    if worker_info is not None:
        # favour using the seed already created for pytorch dataloader workers if it exists
        seed = worker_info.seed
        if increment:
            # space out seed increments so they can't overlap across workers in different iterations
            seed += increment * max(1, worker_info.num_workers)
        return seed
    # fallback to wds rank based seed
    return wds.utils.pytorch_worker_seed()


_SHARD_SHUFFLE_SIZE = 2000
_SHARD_SHUFFLE_INITIAL = 500
_SAMPLE_SHUFFLE_SIZE = 5000
_SAMPLE_SHUFFLE_INITIAL = 1000


class detshuffle2(wds.PipelineStage):
    def __init__(
            self,
            bufsize=1000,
            initial=100,
            seed=0,
            epoch=-1,
    ):
        self.bufsize = bufsize
        self.initial = initial
        self.seed = seed
        self.epoch = epoch

    def run(self, src):
        if isinstance(self.epoch, SharedEpoch):
            epoch = self.epoch.get_value()
        else:
            # NOTE: this is epoch tracking is problematic in a multiprocess (dataloader workers or train)
            # situation as different workers may wrap at different times (or not at all).
            self.epoch += 1
            epoch = self.epoch
        rng = random.Random()
        if self.seed < 0:
            # If seed is negative, we use the worker's seed, this will be different across all nodes/workers
            seed = pytorch_worker_seed(epoch)
        else:
            # This seed to be deterministic AND the same across all nodes/workers in each epoch
            seed = self.seed + epoch
        rng.seed(seed)
        return _shuffle(src, self.bufsize, self.initial, rng)


class ResampledShards2(IterableDataset):
    """An iterable dataset yielding a list of urls."""

    def __init__(
        self,
        urls,
        weights=None,
        nshards=sys.maxsize,
        worker_seed=None,
        deterministic=False,
        epoch=-1,
    ):
        """Sample shards from the shard list with replacement.

        :param urls: a list of URLs as a Python list or brace notation string
        """
        super().__init__()
        urls, weights = expand_urls(urls, weights)
        self.urls = urls
        self.weights = weights
        if self.weights is not None:
            assert len(self.urls) == len(self.weights),\
                f"Number of urls {len(self.urls)} and weights {len(self.weights)} should match."
        assert isinstance(self.urls[0], str)
        self.nshards = nshards
        self.rng = random.Random()
        self.worker_seed = worker_seed
        self.deterministic = deterministic
        self.epoch = epoch

    def __iter__(self):
        """Return an iterator over the shards."""
        if isinstance(self.epoch, SharedEpoch):
            epoch = self.epoch.get_value()
        else:
            # NOTE: this is epoch tracking is problematic in a multiprocess (dataloader workers or train)
            # situation as different workers may wrap at different times (or not at all).
            self.epoch += 1
            epoch = self.epoch
        if self.deterministic:
            # reset seed w/ epoch if deterministic
            if self.worker_seed is None:
                # pytorch worker seed should be deterministic due to being init by arg.seed + rank + worker id
                seed = pytorch_worker_seed(epoch)
            else:
                seed = self.worker_seed() + epoch
            self.rng.seed(seed)
        for _ in range(self.nshards):
            if self.weights is None:
                yield dict(url=self.rng.choice(self.urls))
            else:
                yield dict(url=self.rng.choices(self.urls, weights=self.weights, k=1)[0])


def image_text_dict_collation_fn(samples):
    """Customize collation_fn to generate dict batch """
    assert isinstance(samples[0], (list, tuple)), type(samples[0])
    batched = list(zip(*samples))
    result = dict()
    import torch
    import numpy as np
    for b in batched:
        b = torch.stack(list(b))
        if b.dim()>=3: # dim means image
            result['img']=b
        else:
            result['text']=b
    
    return result

def decode_image(png_bytes):
    return Image.open(BytesIO(png_bytes))


def process_sample(sample):

    if "png" not in sample:
        sample["png"] = b''
    else:
        sample["png"] = decode_image(sample["png"])
    
    sample = {"png": sample["png"], "json": sample["json"]}
    return sample

def get_wds_data(args, is_train, epoch=0, floor=False, wds_processor=None):
    if args.data_path and (args.train_data_weights is None):
         args.train_data_weights = [1.0] * len(args.data_path)
         
    input_shards = args.data_path if is_train else args.valid_data_path
    input_weights = args.train_data_weights if is_train else args.valid_data_weights
    
    assert input_shards is not None

    resampled = getattr(args, 'dataset_resampled', False)

    num_shards = None
    if is_train:
        if args.train_num_samples is not None:
            num_samples = args.train_num_samples
        else:
            num_samples, num_shards = get_dataset_size(input_shards)
            if not num_samples:
                raise RuntimeError(
                    'Currently, the number of dataset samples must be specified for the training dataset. '
                    'Please specify it via `--train-num-samples` if no dataset length info is present.')
    else:
        # Eval will just exhaust the iterator if the size is not specified.
        num_samples = args.val_num_samples or 0 
    weights, weighted_num_samples = get_normalized_weights_and_num_samples(input_weights, num_samples)
    shared_epoch = SharedEpoch(epoch=epoch)  # create a shared epoch store to sync epoch to dataloader worker proc
    
    if resampled:
        complete_url_list = []
        complete_weights = []
        for i, (urls, weights) in enumerate(zip(input_shards, weights)):
            current_url_list = expand_urls(urls)[0]
            complete_url_list.extend(current_url_list)
            per_url_weight = weights / len(current_url_list)
            complete_weights.extend([per_url_weight] * len(current_url_list))
        # pipeline = [ResampledShards2(
        #     complete_url_list,
        #     weights=complete_weights,
        #     deterministic=True,
        #     epoch=shared_epoch,
        # )]
        pipeline = [wds.SimpleShardList(complete_url_list)]
    else:
        # assert args.train_data_upsampling_factors is None,\
        #    "--train_data_upsampling_factors is only supported when sampling with replacement (with --dataset-resampled)."
        pipeline = [wds.SimpleShardList(input_shards)]

    # at this point we have an iterator over all the shards
    if is_train:
        if not resampled:
            pipeline.extend([
                detshuffle2(
                    bufsize=_SHARD_SHUFFLE_SIZE,
                    initial=_SHARD_SHUFFLE_INITIAL,
                    # seed=args.seed,
                    epoch=shared_epoch,
                ),
                wds.split_by_node,
                wds.split_by_worker,
            ])
        pipeline.extend([
            # at this point, we have an iterator over the shards assigned to each worker at each node
            # tarfile_to_samples_nothrow,  # wds.tarfile_to_samples(handler=log_and_continue),
            wds.tarfile_to_samples(handler=log_and_continue),
            wds.shuffle(
                bufsize=_SAMPLE_SHUFFLE_SIZE,
                initial=_SAMPLE_SHUFFLE_INITIAL,
            ),
        ])
    else:
        pipeline.extend([
            wds.split_by_worker,
            # at this point, we have an iterator over the shards assigned to each worker
            wds.tarfile_to_samples(handler=log_and_continue),
        ])
    ### build preprocess_img and preprocess_text from args
    # from .transforms import get_clip_transforms
    # preprocess_img = get_clip_transforms(image_size=data_args.image_processor.crop_size)
    
    # assert (
    #     args.tokenizer.name in ['HFGPT2Tokenizer','HFGPT2TokenizerFast','HFTokenizer']
    #     ), f"Webdataset only support HFTokenizer, HFGPT2Tokenizer or HFGPT2TokenizerFast"
    
    # tokenize = args.tokenizer.tokenize
    
        
    pipeline.extend([
        # wds.select(filter_no_caption_or_no_image),
        wds.decode("pilrgb", handler=log_and_continue),
        wds.rename(image="jpg;png;jpeg;webp", text="json"),
        wds.to_tuple("image", "text"),
        wds.map(wds_processor)
        # wds.map_dict(image=preprocess_img,  text=lambda text: tokenize(text)[0]),
       
        # wds.batched(args.batch_size, collation_fn=image_text_dict_collation_fn, partial=not is_train)
    ])
    # pipeline.extend([
    #     wds.map(process_sample),
    #     wds.rename(image="jpg;png;jpeg;webp", text="json"),
    #     wds.to_tuple("image", "text")
    #     wds.map(wds_processor)
    #     ])

    dataset = wds.DataPipeline(*pipeline)

    if is_train:
        if not resampled:
            num_shards = num_shards or len(expand_urls(input_shards)[0])
            # assert num_shards >= args.num_workers * args.world_size, 'number of shards must be >= total workers'
        # roll over and repeat a few samples to get same number of full batches on each node
        round_fn = math.floor if floor else math.ceil
        global_batch_size = args.batch_size * args.world_size
        num_batches = round_fn(num_samples / global_batch_size)
        num_workers = max(1, args.dataloader_num_workers)
        num_worker_batches = round_fn(num_batches / num_workers)  # per dataloader worker
        num_batches = num_worker_batches * num_workers
        num_samples = num_batches * global_batch_size
        dataset = dataset.with_epoch(num_worker_batches)  # each worker is iterating over this
        # dataset = dataset.with_epoch(num_samples)
    else:
        # last batches are partial, eval is done on single (master) node
        num_batches = math.ceil(num_samples / args.batch_size)

    # dataloader = wds.WebLoader(
    #     dataset,
    #     batch_size=None,
    #     shuffle=False,
    #     num_workers=args.num_workers,
    #     persistent_workers=not (args.num_workers == 0),  # set persistent_workers to false if num_workers is 0
    # )

    # FIXME not clear which approach is better, with_epoch before vs after dataloader?
    # hoping to resolve via https://github.com/webdataset/webdataset/issues/169
    # if is_train:
    #     # roll over and repeat a few samples to get same number of full batches on each node
    #     global_batch_size = args.batch_size * args.world_size
    #     num_batches = math.ceil(num_samples / global_batch_size)
    #     num_workers = max(1, args.num_workers)
    #     num_batches = math.ceil(num_batches / num_workers) * num_workers
    #     num_samples = num_batches * global_batch_size
    #     dataloader = dataloader.with_epoch(num_batches)
    # else:
    #     # last batches are partial, eval is done on single (master) node
    #     num_batches = math.ceil(num_samples / args.batch_size)

    # add meta-data to dataloader instance for convenience
    # dataloader.num_batches = num_batches
    # dataloader.num_samples = num_samples

    return dataset


# def get_data(args, preprocess_fns, epoch=0, tokenizer=None):
#     preprocess_train, preprocess_val = preprocess_fns
#     data = {}

#     if args.train_data or args.dataset_type == "synthetic":
#         data["train"] = get_dataset_fn(args.train_data, args.dataset_type)(
#             args, preprocess_train, is_train=True, epoch=epoch, tokenizer=tokenizer)

#     if args.val_data:
#         data["val"] = get_dataset_fn(args.val_data, args.dataset_type)(
#             args, preprocess_val, is_train=False, tokenizer=tokenizer)

#     return data