lattifai 0.4.5__py3-none-any.whl → 1.0.0__py3-none-any.whl

lattifai/{tokenizer → alignment}/tokenizer.py

@@ -1,27 +1,118 @@
 import gzip
-import inspect
 import pickle
 import re
 from collections import defaultdict
-from typing import Any, Dict, List, Optional, Tuple, Type, TypeVar, Union
+from typing import Any, Dict, List, Optional, Tuple, Type, TypeVar
 
 import torch
 
-from lattifai.errors import LATTICE_DECODING_FAILURE_HELP, LatticeDecodingError
-from lattifai.io import Supervision
-from lattifai.tokenizer.phonemizer import G2Phonemizer
+from lattifai.alignment.phonemizer import G2Phonemizer
+from lattifai.caption import Supervision
+from lattifai.caption import normalize_text as normalize_html_text
+from lattifai.errors import (
+    LATTICE_DECODING_FAILURE_HELP,
+    LatticeDecodingError,
+    ModelLoadError,
+    QuotaExceededError,
+)
 
 PUNCTUATION = '!"#$%&()*+,-./:;<=>?@[\\]^_`{|}~'
 END_PUNCTUATION = '.!?"]。！？”】'
-PUNCTUATION_SPACE = PUNCTUATION + ' '
-STAR_TOKEN = '※'
+PUNCTUATION_SPACE = PUNCTUATION + " "
+STAR_TOKEN = "※"
 
-GROUPING_SEPARATOR = '✹'
+GROUPING_SEPARATOR = "✹"
 
 MAXIMUM_WORD_LENGTH = 40
 
 
-TokenizerT = TypeVar('TokenizerT', bound='LatticeTokenizer')
+TokenizerT = TypeVar("TokenizerT", bound="LatticeTokenizer")
+
+
+def _is_punctuation(char: str) -> bool:
+    """Check if a character is punctuation (not space, not alphanumeric, not CJK)."""
+    if len(char) != 1:
+        return False
+    if char.isspace():
+        return False
+    if char.isalnum():
+        return False
+    # Check if it's a CJK character
+    if "\u4e00" <= char <= "\u9fff":
+        return False
+    # Check if it's an accented Latin character
+    if "\u00c0" <= char <= "\u024f":
+        return False
+    return True
+
+
+def tokenize_multilingual_text(text: str, keep_spaces: bool = True, attach_punctuation: bool = False) -> list[str]:
+    """
+    Tokenize a mixed Chinese-English string into individual units.
+
+    Tokenization rules:
+    - Chinese characters (CJK) are split individually
+    - Consecutive Latin letters (including accented characters) and digits are grouped as one unit
+    - English contractions ('s, 't, 'm, 'll, 're, 've) are kept with the preceding word
+    - Other characters (punctuation, spaces) are split individually by default
+    - If attach_punctuation=True, punctuation marks are attached to the preceding token
+
+    Args:
+        text: Input string containing mixed Chinese and English text
+        keep_spaces: If True, spaces are included in the output as separate tokens.
+                     If False, spaces are excluded from the output. Default is True.
+        attach_punctuation: If True, punctuation marks are attached to the preceding token.
+                            For example, "Hello, World!" becomes ["Hello,", " ", "World!"].
+                            Default is False.
+
+    Returns:
+        List of tokenized units
+
+    Examples:
+        >>> tokenize_multilingual_text("Hello世界")
+        ['Hello', '世', '界']
+        >>> tokenize_multilingual_text("I'm fine")
+        ["I'm", ' ', 'fine']
+        >>> tokenize_multilingual_text("I'm fine", keep_spaces=False)
+        ["I'm", 'fine']
+        >>> tokenize_multilingual_text("Kühlschrank")
+        ['Kühlschrank']
+        >>> tokenize_multilingual_text("Hello, World!", attach_punctuation=True)
+        ['Hello,', ' ', 'World!']
+    """
+    # Regex pattern:
+    # - [a-zA-Z0-9\u00C0-\u024F]+ matches Latin letters (including accented chars like ü, ö, ä, ß, é, etc.)
+    # - (?:'[a-zA-Z]{1,2})? optionally matches contractions like 's, 't, 'm, 'll, 're, 've
+    # - [\u4e00-\u9fff] matches CJK characters
+    # - . matches any other single character
+    # Unicode ranges:
+    # - \u00C0-\u00FF: Latin-1 Supplement (À-ÿ)
+    # - \u0100-\u017F: Latin Extended-A
+    # - \u0180-\u024F: Latin Extended-B
+    pattern = re.compile(r"([a-zA-Z0-9\u00C0-\u024F]+(?:'[a-zA-Z]{1,2})?|[\u4e00-\u9fff]|.)")
+
+    # filter(None, ...) removes any empty strings from re.findall results
+    tokens = list(filter(None, pattern.findall(text)))
+
+    if attach_punctuation and len(tokens) > 1:
+        # Attach punctuation to the preceding token
+        # Punctuation characters (excluding spaces) are merged with the previous token
+        merged_tokens = []
+        i = 0
+        while i < len(tokens):
+            token = tokens[i]
+            # Look ahead to collect consecutive punctuation (non-space, non-alphanumeric, non-CJK)
+            if merged_tokens and _is_punctuation(token):
+                merged_tokens[-1] = merged_tokens[-1] + token
+            else:
+                merged_tokens.append(token)
+            i += 1
+        tokens = merged_tokens
+
+    if not keep_spaces:
+        tokens = [t for t in tokens if not t.isspace()]
+
+    return tokens
 
 
 class LatticeTokenizer:
@@ -29,12 +120,13 @@ class LatticeTokenizer:
 
     def __init__(self, client_wrapper: Any):
         self.client_wrapper = client_wrapper
+        self.model_name = ""
         self.words: List[str] = []
         self.g2p_model: Any = None  # Placeholder for G2P model
         self.dictionaries = defaultdict(lambda: [])
-        self.oov_word = '<unk>'
+        self.oov_word = "<unk>"
         self.sentence_splitter = None
-        self.device = 'cpu'
+        self.device = "cpu"
 
     def init_sentence_splitter(self):
         if self.sentence_splitter is not None:
@@ -45,14 +137,14 @@ class LatticeTokenizer:
 
         providers = []
         device = self.device
-        if device.startswith('cuda') and ort.get_all_providers().count('CUDAExecutionProvider') > 0:
-            providers.append('CUDAExecutionProvider')
-        elif device.startswith('mps') and ort.get_all_providers().count('MPSExecutionProvider') > 0:
-            providers.append('MPSExecutionProvider')
+        if device.startswith("cuda") and ort.get_all_providers().count("CUDAExecutionProvider") > 0:
+            providers.append("CUDAExecutionProvider")
+        elif device.startswith("mps") and ort.get_all_providers().count("MPSExecutionProvider") > 0:
+            providers.append("MPSExecutionProvider")
 
         sat = SaT(
-            'sat-3l-sm',
-            ort_providers=providers + ['CPUExecutionProvider'],
+            "sat-3l-sm",
+            ort_providers=providers + ["CPUExecutionProvider"],
         )
         self.sentence_splitter = sat
 
@@ -79,23 +171,23 @@ class LatticeTokenizer:
         # or other forms like [SOMETHING] SPEAKER:
 
         # Pattern 1: [mark] HTML-encoded separator speaker:
-        pattern1 = r'^(\[[^\]]+\])\s+(&gt;&gt;|>>)\s+(.+)$'
+        pattern1 = r"^(\[[^\]]+\])\s+(&gt;&gt;|>>)\s+(.+)$"
         match1 = re.match(pattern1, sentence.strip())
         if match1:
             special_mark = match1.group(1)
             separator = match1.group(2)
             speaker_part = match1.group(3)
-            return [special_mark, f'{separator} {speaker_part}']
+            return [special_mark, f"{separator} {speaker_part}"]
 
         # Pattern 2: [mark] speaker:
-        pattern2 = r'^(\[[^\]]+\])\s+([^:]+:)(.*)$'
+        pattern2 = r"^(\[[^\]]+\])\s+([^:]+:)(.*)$"
         match2 = re.match(pattern2, sentence.strip())
         if match2:
             special_mark = match2.group(1)
             speaker_label = match2.group(2)
             remaining = match2.group(3).strip()
             if remaining:
-                return [special_mark, f'{speaker_label} {remaining}']
+                return [special_mark, f"{speaker_label} {remaining}"]
             else:
                 return [special_mark, speaker_label]
 
@@ -107,28 +199,45 @@ class LatticeTokenizer:
         cls: Type[TokenizerT],
         client_wrapper: Any,
         model_path: str,
-        device: str = 'cpu',
+        model_name: str,
+        device: str = "cpu",
         compressed: bool = True,
     ) -> TokenizerT:
         """Load tokenizer from exported binary file"""
         from pathlib import Path
 
-        words_model_path = f'{model_path}/words.bin'
-        if compressed:
-            with gzip.open(words_model_path, 'rb') as f:
-                data = pickle.load(f)
-        else:
-            with open(words_model_path, 'rb') as f:
-                data = pickle.load(f)
+        words_model_path = f"{model_path}/words.bin"
+        try:
+            if compressed:
+                with gzip.open(words_model_path, "rb") as f:
+                    data = pickle.load(f)
+            else:
+                with open(words_model_path, "rb") as f:
+                    data = pickle.load(f)
+        except pickle.UnpicklingError as e:
+            del e
+            import msgpack
+
+            if compressed:
+                with gzip.open(words_model_path, "rb") as f:
+                    data = msgpack.unpack(f, raw=False, strict_map_key=False)
+            else:
+                with open(words_model_path, "rb") as f:
+                    data = msgpack.unpack(f, raw=False, strict_map_key=False)
 
         tokenizer = cls(client_wrapper=client_wrapper)
-        tokenizer.words = data['words']
-        tokenizer.dictionaries = defaultdict(list, data['dictionaries'])
-        tokenizer.oov_word = data['oov_word']
-
-        g2p_model_path = f'{model_path}/g2p.bin' if Path(f'{model_path}/g2p.bin').exists() else None
-        if g2p_model_path:
-            tokenizer.g2p_model = G2Phonemizer(g2p_model_path, device=device)
+        tokenizer.model_name = model_name
+        tokenizer.words = data["words"]
+        tokenizer.dictionaries = defaultdict(list, data["dictionaries"])
+        tokenizer.oov_word = data["oov_word"]
+
+        g2pp_model_path = f"{model_path}/g2pp.bin" if Path(f"{model_path}/g2pp.bin").exists() else None
+        if g2pp_model_path:
+            tokenizer.g2p_model = G2Phonemizer(g2pp_model_path, device=device)
+        else:
+            g2p_model_path = f"{model_path}/g2p.bin" if Path(f"{model_path}/g2p.bin").exists() else None
+            if g2p_model_path:
+                tokenizer.g2p_model = G2Phonemizer(g2p_model_path, device=device)
 
         tokenizer.device = device
         tokenizer.add_special_tokens()
@@ -136,18 +245,22 @@ class LatticeTokenizer:
 
     def add_special_tokens(self):
         tokenizer = self
-        for special_token in ['&gt;&gt;', '&gt;']:
+        for special_token in ["&gt;&gt;", "&gt;"]:
             if special_token not in tokenizer.dictionaries:
                 tokenizer.dictionaries[special_token] = tokenizer.dictionaries[tokenizer.oov_word]
         return self
 
     def prenormalize(self, texts: List[str], language: Optional[str] = None) -> List[str]:
         if not self.g2p_model:
-            raise ValueError('G2P model is not loaded, cannot prenormalize texts')
+            raise ValueError("G2P model is not loaded, cannot prenormalize texts")
 
         oov_words = []
         for text in texts:
-            words = text.lower().replace('-', ' ').replace('—', ' ').replace('–', ' ').split()
+            text = normalize_html_text(text)
+            # support english, chinese and german tokenization
+            words = tokenize_multilingual_text(
+                text.lower().replace("-", " ").replace("—", " ").replace("–", " "), keep_spaces=False
+            )
             oovs = [w.strip(PUNCTUATION) for w in words if w not in self.words]
             if oovs:
                 oov_words.extend([w for w in oovs if (w not in self.words and len(w) <= MAXIMUM_WORD_LENGTH)])
@@ -156,7 +269,7 @@ class LatticeTokenizer:
         if oov_words:
             indexs = []
             for k, _word in enumerate(oov_words):
-                if any(_word.startswith(p) and _word.endswith(q) for (p, q) in [('(', ')'), ('[', ']')]):
+                if any(_word.startswith(p) and _word.endswith(q) for (p, q) in [("(", ")"), ("[", "]")]):
                     self.dictionaries[_word] = self.dictionaries[self.oov_word]
                 else:
                     _word = _word.strip(PUNCTUATION_SPACE)
@@ -187,38 +300,49 @@ class LatticeTokenizer:
 
         Carefull about speaker changes.
         """
-        texts, text_len, sidx = [], 0, 0
-        speakers = []
+        texts, speakers = [], []
+        text_len, sidx = 0, 0
+
+        def flush_segment(end_idx: int, speaker: Optional[str] = None):
+            """Flush accumulated text from sidx to end_idx with given speaker."""
+            nonlocal text_len, sidx
+            if sidx <= end_idx:
+                if len(speakers) < len(texts) + 1:
+                    speakers.append(speaker)
+                text = " ".join(sup.text for sup in supervisions[sidx : end_idx + 1])
+                texts.append(text)
+                sidx = end_idx + 1
+                text_len = 0
+
         for s, supervision in enumerate(supervisions):
             text_len += len(supervision.text)
+            is_last = s == len(supervisions) - 1
+
             if supervision.speaker:
+                # Flush previous segment without speaker (if any)
                 if sidx < s:
-                    if len(speakers) < len(texts) + 1:
-                        speakers.append(None)
-                    text = ' '.join([sup.text for sup in supervisions[sidx:s]])
-                    texts.append(text)
-                    sidx = s
+                    flush_segment(s - 1, None)
                     text_len = len(supervision.text)
-                speakers.append(supervision.speaker)
 
-            else:
-                if text_len >= 2000 or s == len(supervisions) - 1:
-                    if len(speakers) < len(texts) + 1:
-                        speakers.append(None)
-                    text = ' '.join([sup.text for sup in supervisions[sidx : s + 1]])
-                    texts.append(text)
-                    sidx = s + 1
-                    text_len = 0
-
-        assert len(speakers) == len(texts), f'len(speakers)={len(speakers)} != len(texts)={len(texts)}'
+                # Check if we should flush this speaker's segment now
+                next_has_speaker = not is_last and supervisions[s + 1].speaker
+                if is_last or next_has_speaker:
+                    flush_segment(s, supervision.speaker)
+                else:
+                    speakers.append(supervision.speaker)
+
+            elif text_len >= 2000 or is_last:
+                flush_segment(s, None)
+
+        assert len(speakers) == len(texts), f"len(speakers)={len(speakers)} != len(texts)={len(texts)}"
         sentences = self.sentence_splitter.split(texts, threshold=0.15, strip_whitespace=strip_whitespace)
 
-        supervisions, remainder = [], ''
+        supervisions, remainder = [], ""
         for k, (_speaker, _sentences) in enumerate(zip(speakers, sentences)):
             # Prepend remainder from previous iteration to the first sentence
             if _sentences and remainder:
                 _sentences[0] = remainder + _sentences[0]
-                remainder = ''
+                remainder = ""
 
             if not _sentences:
                 continue
@@ -228,14 +352,14 @@ class LatticeTokenizer:
             for s, _sentence in enumerate(_sentences):
                 if remainder:
                     _sentence = remainder + _sentence
-                    remainder = ''
+                    remainder = ""
                 # Detect and split special sentence types: e.g., '[APPLAUSE] &gt;&gt; MIRA MURATI:' -> ['[APPLAUSE]', '&gt;&gt; MIRA MURATI:']  # noqa: E501
                 resplit_parts = self._resplit_special_sentence_types(_sentence)
-                if any(resplit_parts[-1].endswith(sp) for sp in [':', '：']):
+                if any(resplit_parts[-1].endswith(sp) for sp in [":", "："]):
                     if s < len(_sentences) - 1:
-                        _sentences[s + 1] = resplit_parts[-1] + ' ' + _sentences[s + 1]
+                        _sentences[s + 1] = resplit_parts[-1] + " " + _sentences[s + 1]
                     else:  # last part
-                        remainder = resplit_parts[-1] + ' '
+                        remainder = resplit_parts[-1] + " "
                     processed_sentences.extend(resplit_parts[:-1])
                 else:
                     processed_sentences.extend(resplit_parts)
@@ -243,7 +367,7 @@ class LatticeTokenizer:
 
             if not _sentences:
                 if remainder:
-                    _sentences, remainder = [remainder.strip()], ''
+                    _sentences, remainder = [remainder.strip()], ""
                 else:
                     continue
 
@@ -257,12 +381,12 @@ class LatticeTokenizer:
                     Supervision(text=text, speaker=(_speaker if s == 0 else None))
                     for s, text in enumerate(_sentences[:-1])
                 )
-                remainder = _sentences[-1] + ' ' + remainder
+                remainder = _sentences[-1] + " " + remainder
                 if k < len(speakers) - 1 and speakers[k + 1] is not None:  # next speaker is set
                     supervisions.append(
                         Supervision(text=remainder.strip(), speaker=_speaker if len(_sentences) == 1 else None)
                     )
-                    remainder = ''
+                    remainder = ""
                 elif len(_sentences) == 1:
                     if k == len(speakers) - 1:
                         pass  # keep _speaker for the last supervision
@@ -285,20 +409,23 @@ class LatticeTokenizer:
 
         pronunciation_dictionaries = self.prenormalize([s.text for s in supervisions])
         response = self.client_wrapper.post(
-            'tokenize',
+            "tokenize",
             json={
-                'supervisions': [s.to_dict() for s in supervisions],
-                'pronunciation_dictionaries': pronunciation_dictionaries,
+                "model_name": self.model_name,
+                "supervisions": [s.to_dict() for s in supervisions],
+                "pronunciation_dictionaries": pronunciation_dictionaries,
             },
         )
+        if response.status_code == 402:
+            raise QuotaExceededError(response.json().get("detail", "Quota exceeded"))
         if response.status_code != 200:
-            raise Exception(f'Failed to tokenize texts: {response.text}')
+            raise Exception(f"Failed to tokenize texts: {response.text}")
         result = response.json()
-        lattice_id = result['id']
+        lattice_id = result["id"]
         return (
             supervisions,
             lattice_id,
-            (result['lattice_graph'], result['final_state'], result.get('acoustic_scale', 1.0)),
+            (result["lattice_graph"], result["final_state"], result.get("acoustic_scale", 1.0)),
         )
 
     def detokenize(
@@ -310,16 +437,17 @@ class LatticeTokenizer:
     ) -> List[Supervision]:
         emission, results, labels, frame_shift, offset, channel = lattice_results  # noqa: F841
         response = self.client_wrapper.post(
-            'detokenize',
+            "detokenize",
             json={
-                'lattice_id': lattice_id,
-                'frame_shift': frame_shift,
-                'results': [t.to_dict() for t in results[0]],
-                'labels': labels[0],
-                'offset': offset,
-                'channel': channel,
-                'return_details': return_details,
-                'destroy_lattice': True,
+                "model_name": self.model_name,
+                "lattice_id": lattice_id,
+                "frame_shift": frame_shift,
+                "results": [t.to_dict() for t in results[0]],
+                "labels": labels[0],
+                "offset": offset,
+                "channel": channel,
+                "return_details": False if return_details is None else return_details,
+                "destroy_lattice": True,
             },
         )
         if response.status_code == 422:
@@ -327,94 +455,20 @@ class LatticeTokenizer:
                 lattice_id,
                 original_error=Exception(LATTICE_DECODING_FAILURE_HELP),
             )
+        if response.status_code == 402:
+            raise QuotaExceededError(response.json().get("detail", "Quota exceeded"))
         if response.status_code != 200:
-            raise Exception(f'Failed to detokenize lattice: {response.text}')
+            raise Exception(f"Failed to detokenize lattice: {response.text}")
 
         result = response.json()
-        if not result.get('success'):
-            raise Exception('Failed to detokenize the alignment results.')
+        if not result.get("success"):
+            raise Exception("Failed to detokenize the alignment results.")
 
-        alignments = [Supervision.from_dict(s) for s in result['supervisions']]
+        alignments = [Supervision.from_dict(s) for s in result["supervisions"]]
 
         if return_details:
             # Add emission confidence scores for segments and word-level alignments
-            _add_confidence_scores(alignments, emission, labels[0], frame_shift)
-
-        alignments = _update_alignments_speaker(supervisions, alignments)
-
-        return alignments
-
-
-class AsyncLatticeTokenizer(LatticeTokenizer):
-    async def _post_async(self, endpoint: str, **kwargs):
-        response = self.client_wrapper.post(endpoint, **kwargs)
-        if inspect.isawaitable(response):
-            return await response
-        return response
-
-    async def tokenize(
-        self, supervisions: List[Supervision], split_sentence: bool = False
-    ) -> Tuple[str, Dict[str, Any]]:
-        if split_sentence:
-            self.init_sentence_splitter()
-            supervisions = self.split_sentences(supervisions)
-
-        pronunciation_dictionaries = self.prenormalize([s.text for s in supervisions])
-        response = await self._post_async(
-            'tokenize',
-            json={
-                'supervisions': [s.to_dict() for s in supervisions],
-                'pronunciation_dictionaries': pronunciation_dictionaries,
-            },
-        )
-        if response.status_code != 200:
-            raise Exception(f'Failed to tokenize texts: {response.text}')
-        result = response.json()
-        lattice_id = result['id']
-        return (
-            supervisions,
-            lattice_id,
-            (result['lattice_graph'], result['final_state'], result.get('acoustic_scale', 1.0)),
-        )
-
-    async def detokenize(
-        self,
-        lattice_id: str,
-        lattice_results: Tuple[torch.Tensor, Any, Any, float, float],
-        supervisions: List[Supervision],
-        return_details: bool = False,
-    ) -> List[Supervision]:
-        emission, results, labels, frame_shift, offset, channel = lattice_results  # noqa: F841
-        response = await self._post_async(
-            'detokenize',
-            json={
-                'lattice_id': lattice_id,
-                'frame_shift': frame_shift,
-                'results': [t.to_dict() for t in results[0]],
-                'labels': labels[0],
-                'offset': offset,
-                'channel': channel,
-                'return_details': return_details,
-                'destroy_lattice': True,
-            },
-        )
-        if response.status_code == 422:
-            raise LatticeDecodingError(
-                lattice_id,
-                original_error=Exception(LATTICE_DECODING_FAILURE_HELP),
-            )
-        if response.status_code != 200:
-            raise Exception(f'Failed to detokenize lattice: {response.text}')
-
-        result = response.json()
-        if not result.get('success'):
-            return Exception('Failed to detokenize the alignment results.')
-
-        alignments = [Supervision.from_dict(s) for s in result['supervisions']]
-
-        if return_details:
-            # Add emission confidence scores for segments and word-level alignments
-            _add_confidence_scores(alignments, emission, labels[0], frame_shift)
+            _add_confidence_scores(alignments, emission, labels[0], frame_shift, offset)
 
         alignments = _update_alignments_speaker(supervisions, alignments)
 
@@ -426,6 +480,7 @@ def _add_confidence_scores(
     emission: torch.Tensor,
     labels: List[int],
     frame_shift: float,
+    offset: float = 0.0,
 ) -> None:
     """
     Add confidence scores to supervisions and their word-level alignments.
@@ -443,8 +498,8 @@ def _add_confidence_scores(
     tokens = torch.tensor(labels, dtype=torch.int64, device=emission.device)
 
     for supervision in supervisions:
-        start_frame = int(supervision.start / frame_shift)
-        end_frame = int(supervision.end / frame_shift)
+        start_frame = int((supervision.start - offset) / frame_shift)
+        end_frame = int((supervision.end - offset) / frame_shift)
 
         # Compute segment-level confidence
         probabilities = emission[0, start_frame:end_frame].softmax(dim=-1)
@@ -453,11 +508,11 @@ def _add_confidence_scores(
         supervision.score = round(1.0 - diffprobs.mean().item(), ndigits=4)
 
         # Compute word-level confidence if alignment exists
-        if hasattr(supervision, 'alignment') and supervision.alignment:
-            words = supervision.alignment.get('word', [])
+        if hasattr(supervision, "alignment") and supervision.alignment:
+            words = supervision.alignment.get("word", [])
             for w, item in enumerate(words):
-                start = int(item.start / frame_shift) - start_frame
-                end = int(item.end / frame_shift) - start_frame
+                start = int((item.start - offset) / frame_shift) - start_frame
+                end = int((item.end - offset) / frame_shift) - start_frame
                 words[w] = item._replace(score=round(1.0 - diffprobs[start:end].mean().item(), ndigits=4))
 
 
@@ -472,3 +527,23 @@ def _update_alignments_speaker(supervisions: List[Supervision], alignments: List
     for supervision, alignment in zip(supervisions, alignments):
         alignment.speaker = supervision.speaker
     return alignments
+
+
+def _load_tokenizer(
+    client_wrapper: Any,
+    model_path: str,
+    model_name: str,
+    device: str,
+    *,
+    tokenizer_cls: Type[LatticeTokenizer] = LatticeTokenizer,
+) -> LatticeTokenizer:
+    """Instantiate tokenizer with consistent error handling."""
+    try:
+        return tokenizer_cls.from_pretrained(
+            client_wrapper=client_wrapper,
+            model_path=model_path,
+            model_name=model_name,
+            device=device,
+        )
+    except Exception as e:
+        raise ModelLoadError(f"tokenizer from {model_path}", original_error=e)