Chunk file reads and tokenization

llmfoundry/data/data.py

@@ -4,6 +4,7 @@
 """Datasets for converting to MDS Shards."""
 import os
 import warnings
+from abc import ABC, abstractmethod
 from typing import Dict, Iterable, Union
 
 import datasets as hf_datasets
@@ -35,39 +36,20 @@ def __iter__(self) -> Iterable[Dict[str, bytes]]:
             yield {'text': sample['text'].encode('utf-8')}
 
 
-class ConcatTokensDataset(IterableDataset):
-    """An IterableDataset that returns token samples for MDSWriter.
-
-    Returns dicts of {'tokens': bytes}
-
-    To use data created by this class and written to MDS format:
-
-    ```python
-        import torch
-        from streaming.base import StreamingDataset
-        from transformers import AutoTokenizer
+class AbstractConcatTokensDataset(ABC, IterableDataset):
+    """Abstract class for defining an IterableDataset that tokenizes and.
 
-        tokenizer = AutoTokenizer.from_pretrained('your/tokenizer')
-        ds = StreamingDataset(local='mds-data-folder', split='val')
-
-        # note, you need to copy the numpy array because the original is non-writeable
-        # and torch does not support non-writeable tensors, so you get a scary warning and
-        # if you do try to write to the tensor you get undefined behavior
-        tokens = torch.from_numpy(np.frombuffer(ds[0]['tokens'], dtype=np.int64).copy())
-        print(tokenizer.decode(tokens))
-    ```
+    concatenates text samples on the fly.
     """
 
     def __init__(
         self,
-        hf_dataset: Union[hf_datasets.IterableDataset, hf_datasets.Dataset],
         tokenizer: PreTrainedTokenizerBase,
         max_length: int,
         bos_text: str,
         eos_text: str,
         no_wrap: bool,
     ):
-        self.hf_dataset = hf_dataset
         self.tokenizer = tokenizer
         os.environ['TOKENIZERS_PARALLELISM'] = 'false'
         self.max_length = max_length
@@ -114,8 +96,47 @@ def __init__(
                 'in duplicated special tokens. Please be sure this is what you intend.',
             )
 
+    @abstractmethod
     def __iter__(self) -> Iterable[Dict[str, bytes]]:
+        pass
+
+
+class ConcatTokensDataset(AbstractConcatTokensDataset):
+    """An IterableDataset that returns token samples for MDSWriter.
+
+    Returns dicts of {'tokens': bytes}
+
+    To use data created by this class and written to MDS format:
 
+    ```python
+        import torch
+        from streaming.base import StreamingDataset
+        from transformers import AutoTokenizer
+
+        tokenizer = AutoTokenizer.from_pretrained('your/tokenizer')
+        ds = StreamingDataset(local='mds-data-folder', split='val')
+
+        # note, you need to copy the numpy array because the original is non-writeable
+        # and torch does not support non-writeable tensors, so you get a scary warning and
+        # if you do try to write to the tensor you get undefined behavior
+        tokens = torch.from_numpy(np.frombuffer(ds[0]['tokens'], dtype=np.int64).copy())
+        print(tokenizer.decode(tokens))
+    ```
+    """
+
+    def __init__(
+        self,
+        hf_dataset: Union[hf_datasets.IterableDataset, hf_datasets.Dataset],
+        tokenizer: PreTrainedTokenizerBase,
+        max_length: int,
+        bos_text: str,
+        eos_text: str,
+        no_wrap: bool,
+    ):
+        self.hf_dataset = hf_dataset
+        super().__init__(tokenizer, max_length, bos_text, eos_text, no_wrap)
+
+    def __iter__(self) -> Iterable[Dict[str, bytes]]:
         buffer = []
         for sample in self.hf_dataset:
             encoded = self.tokenizer(

llmfoundry/utils/data_prep_utils.py

@@ -108,7 +108,7 @@ def __init__(
         output_folder: str,
         object_store: Optional[ObjectStore],
     ):
-        """Iterable that downloads files from an object store before yielding.
+        """Iterable that downloads files before yielding the local filename.
 
         If object_store is None, input_folder_prefix is treated as a local path.
 
@@ -138,7 +138,4 @@ def __iter__(self):
                     object_name=object_name,
                     output_filename=output_filename,
                 )
-
-            with open(output_filename) as _txt_file:
-                txt = _txt_file.read()
-            yield {'text': txt}
+            yield output_filename

scripts/data_prep/convert_text_to_mds.py

@@ -7,9 +7,11 @@
 import tempfile
 from argparse import ArgumentParser, Namespace
 from concurrent.futures import ProcessPoolExecutor
+from functools import partial
 from glob import glob
-from typing import Iterable, List, Tuple, cast
+from typing import Dict, Iterable, List, Tuple, cast
 
+import numpy as np
 import psutil
 from composer.utils import (
     ObjectStore,
@@ -18,9 +20,9 @@
 )
 from streaming import MDSWriter
 from tqdm import tqdm
-from transformers import AutoTokenizer
+from transformers import AutoTokenizer, PreTrainedTokenizerBase
 
-from llmfoundry.data import ConcatTokensDataset
+from llmfoundry.data.data import AbstractConcatTokensDataset
 from llmfoundry.utils import maybe_create_mosaicml_logger
 from llmfoundry.utils.data_prep_utils import (
     DownloadingIterable,
@@ -37,6 +39,62 @@
 DONE_FILENAME = '.text_to_mds_conversion_done'
 
 
+class ConcatTokensFromFilesDataset(AbstractConcatTokensDataset):
+    """An IterableDataset that returns token samples for MDSWriter from files.
+
+    Returns dicts of {'tokens': bytes}
+
+    Each file is considered a sequence.
+    """
+
+    def __init__(
+        self,
+        files: Iterable[str],
+        tokenizer: PreTrainedTokenizerBase,
+        max_length: int,
+        bos_text: str,
+        eos_text: str,
+        no_wrap: bool,
+    ):
+        self.files = files
+        super().__init__(tokenizer, max_length, bos_text, eos_text, no_wrap)
+
+    def __iter__(self) -> Iterable[Dict[str, bytes]]:
+
+        buffer = []
+        for file in self.files:
+            with open(file, 'r') as f:
+                buffer += self.bos_tokens
+                first_chunk = True
+                # Read the file in 1MB chunks to avoid memory issues
+                for chunk in iter(partial(f.read, 1000000), ''):
+                    # Tokenize the chunk
+                    encoded = self.tokenizer(
+                        chunk,
+                        truncation=False,
+                        padding=False,
+                    )
+                    iids = encoded['input_ids']
+
+                    # If this is not the first chunk, remove the BOS token
+                    if not first_chunk:
+                        if iids[0] == self.tokenizer.bos_token_id:
+                            iids = iids[1:]
+
+                    # Add the tokens to the buffer
+                    buffer += iids
+                    while len(buffer) >= self.max_length:
+                        concat_sample = buffer[:self.max_length]
+                        buffer = buffer[self.
+                                        max_length:] if self.should_wrap else []
+                        yield {'tokens': np.asarray(concat_sample).tobytes()}
+
+                    first_chunk = False
+
+                # Add the EOS token to the buffer to separate files.
+                buffer += self.eos_tokens
+
+
 def parse_args() -> Namespace:
     """Parse commandline arguments."""
     parser = ArgumentParser(
@@ -277,8 +335,8 @@ def download_and_convert(
 
         # Use the ConcatTokensDataset from LLM-foundry to concatenate sequences of tokens up
         # to the maximum sequence length
-        dataset = ConcatTokensDataset(
-            hf_dataset=downloading_iter,
+        dataset = ConcatTokensFromFilesDataset(
+            files=downloading_iter,
             max_length=concat_tokens,
             tokenizer=tokenizer,
             eos_text=eos_text,