Release 20230314

Co-authored-by: Akash Mahajan <akash.mahajan@microsoft.com>
Co-authored-by: Jong Wook Kim <jongwook@openai.com>

.gitattributes

@@ -0,0 +1,3 @@
+# Override jupyter in Github language stats for more accurate estimate of repo code languages
+# reference: https://github.com/github/linguist/blob/master/docs/overrides.md#generated-code
+*.ipynb linguist-generated

CHANGELOG.md

@@ -1,6 +1,20 @@
 # CHANGELOG
 
-## [v20230307](https://github.com/openai/whisper/releases/tag/v202303067)
+## [v20230314](https://github.com/openai/whisper/releases/tag/v20230314)
+
+* abort find_alignment on empty input ([#1090](https://github.com/openai/whisper/pull/1090))
+* Fix truncated words list when the replacement character is decoded ([#1089](https://github.com/openai/whisper/pull/1089))
+* fix github language stats getting dominated by jupyter notebook ([#1076](https://github.com/openai/whisper/pull/1076))
+* Fix alignment between the segments and the list of words ([#1087](https://github.com/openai/whisper/pull/1087))
+* Use tiktoken ([#1044](https://github.com/openai/whisper/pull/1044))
+
+## [v20230308](https://github.com/openai/whisper/releases/tag/v20230308)
+
+* kwargs in decode() for convenience ([#1061](https://github.com/openai/whisper/pull/1061))
+* fix all_tokens handling that caused more repetitions and discrepancy in JSON ([#1060](https://github.com/openai/whisper/pull/1060))
+* fix typo in CHANGELOG.md
+
+## [v20230307](https://github.com/openai/whisper/releases/tag/v20230307)
 
 * Fix the repetition/hallucination issue identified in #1046 ([#1052](https://github.com/openai/whisper/pull/1052))
 * Use triton==2.0.0 ([#1053](https://github.com/openai/whisper/pull/1053))

MANIFEST.in

@@ -2,6 +2,4 @@ include requirements.txt
 include README.md
 include LICENSE
 include whisper/assets/*
-include whisper/assets/gpt2/*
-include whisper/assets/multilingual/*
 include whisper/normalizers/english.json

requirements.txt

@@ -3,5 +3,5 @@ numpy
 torch
 tqdm
 more-itertools
-transformers>=4.19.0
+tiktoken==0.3.1
 ffmpeg-python==0.2.0

tests/test_tokenizer.py

@@ -12,3 +12,13 @@ def test_tokenizer():
     assert gpt2_tokenizer.decode(gpt2_tokens) == text
     assert multilingual_tokenizer.decode(multilingual_tokens) == text
     assert len(gpt2_tokens) > len(multilingual_tokens)
+
+
+def test_split_on_unicode():
+    multilingual_tokenizer = get_tokenizer(multilingual=True)
+
+    tokens = [8404, 871, 287, 6, 246, 526, 3210, 20378]
+    words, word_tokens = multilingual_tokenizer.split_tokens_on_unicode(tokens)
+
+    assert words == [" elle", " est", " l", "'", "�", "é", "rit", "oire"]
+    assert word_tokens == [[8404], [871], [287], [6], [246], [526], [3210], [20378]]

tests/test_transcribe.py

@@ -4,6 +4,7 @@ import pytest
 import torch
 
 import whisper
+from whisper.tokenizer import get_tokenizer
 
 
 @pytest.mark.parametrize("model_name", whisper.available_models())
@@ -17,12 +18,18 @@ def test_transcribe(model_name: str):
         audio_path, language=language, temperature=0.0, word_timestamps=True
     )
     assert result["language"] == "en"
+    assert result["text"] == "".join([s["text"] for s in result["segments"]])
 
     transcription = result["text"].lower()
     assert "my fellow americans" in transcription
     assert "your country" in transcription
     assert "do for you" in transcription
 
+    tokenizer = get_tokenizer(model.is_multilingual)
+    all_tokens = [t for s in result["segments"] for t in s["tokens"]]
+    assert tokenizer.decode(all_tokens) == result["text"]
+    assert tokenizer.decode_with_timestamps(all_tokens).startswith("<|0.00|>")
+
     timing_checked = False
     for segment in result["segments"]:
         for timing in segment["words"]:
@@ -30,7 +37,6 @@ def test_transcribe(model_name: str):
             if timing["word"].strip(" ,") == "Americans":
                 assert timing["start"] <= 1.8
                 assert timing["end"] >= 1.8
-                print(timing)
                 timing_checked = True
 
     assert timing_checked

whisper/assets/gpt2/special_tokens_map.json

@@ -1 +0,0 @@
-{"bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|endoftext|>"}

whisper/assets/gpt2/tokenizer_config.json

@@ -1 +0,0 @@
-{"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "gpt2", "tokenizer_class": "GPT2Tokenizer"}

whisper/assets/multilingual/added_tokens.json

@@ -1 +0,0 @@
-{"<|endoftext|>": 50257}

whisper/assets/multilingual/special_tokens_map.json

@@ -1 +0,0 @@
-{"bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|endoftext|>"}

whisper/assets/multilingual/tokenizer_config.json

@@ -1 +0,0 @@
-{"unk_token": {"content": "<|endoftext|>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "bos_token": {"content": "<|endoftext|>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "eos_token": {"content": "<|endoftext|>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "multilingual", "errors": "replace", "tokenizer_class": "GPT2Tokenizer"}

whisper/decoding.py

@@ -1,4 +1,4 @@
-from dataclasses import dataclass, field
+from dataclasses import dataclass, field, replace
 from typing import TYPE_CHECKING, Dict, Iterable, List, Optional, Sequence, Tuple, Union
 
 import numpy as np
@@ -778,7 +778,10 @@ class DecodingTask:
 
 @torch.no_grad()
 def decode(
-    model: "Whisper", mel: Tensor, options: DecodingOptions = DecodingOptions()
+    model: "Whisper",
+    mel: Tensor,
+    options: DecodingOptions = DecodingOptions(),
+    **kwargs,
 ) -> Union[DecodingResult, List[DecodingResult]]:
     """
     Performs decoding of 30-second audio segment(s), provided as Mel spectrogram(s).
@@ -802,6 +805,9 @@ def decode(
     if single := mel.ndim == 2:
         mel = mel.unsqueeze(0)
 
+    if kwargs:
+        options = replace(options, **kwargs)
+
     result = DecodingTask(model, options).run(mel)
 
     return result[0] if single else result

whisper/timing.py

@@ -1,3 +1,4 @@
+import itertools
 import subprocess
 import warnings
 from dataclasses import dataclass
@@ -169,6 +170,9 @@ def find_alignment(
     medfilt_width: int = 7,
     qk_scale: float = 1.0,
 ) -> List[WordTiming]:
+    if len(text_tokens) == 0:
+        return []
+
     tokens = torch.tensor(
         [
             *tokenizer.sot_sequence,
@@ -290,34 +294,41 @@ def add_word_timestamps(
     if len(segments) == 0:
         return
 
-    text_tokens = [t for segment in segments for t in segment["tokens"]]
+    text_tokens_per_segment = [
+        [token for token in segment["tokens"] if token < tokenizer.eot]
+        for segment in segments
+    ]
+
+    text_tokens = list(itertools.chain.from_iterable(text_tokens_per_segment))
     alignment = find_alignment(model, tokenizer, text_tokens, mel, num_frames, **kwargs)
     merge_punctuations(alignment, prepend_punctuations, append_punctuations)
 
     time_offset = segments[0]["seek"] * HOP_LENGTH / SAMPLE_RATE
-    segment_lengths = [len(s["tokens"]) for s in segments]
-    token_sources = np.repeat(np.arange(len(segments)), segment_lengths)
+    word_index = 0
 
-    for segment in segments:
-        segment["words"] = []
+    for segment, text_tokens in zip(segments, text_tokens_per_segment):
+        saved_tokens = 0
+        words = []
 
-    word_boundaries = np.pad(np.cumsum([len(w.tokens) for w in alignment]), (1, 0))
-    for i, timing in enumerate(alignment):
-        if timing.word:
-            segment = segments[token_sources[word_boundaries[i]]]
-            start = round(time_offset + timing.start, 2)
-            end = round(time_offset + timing.end, 2)
-            segment["words"].append(
-                dict(
-                    word=timing.word,
-                    start=start,
-                    end=end,
-                    probability=timing.probability,
+        while word_index < len(alignment) and saved_tokens < len(text_tokens):
+            timing = alignment[word_index]
+
+            if timing.word:
+                words.append(
+                    dict(
+                        word=timing.word,
+                        start=round(time_offset + timing.start, 2),
+                        end=round(time_offset + timing.end, 2),
+                        probability=timing.probability,
+                    )
                 )
-            )
 
-    for segment in segments:
-        if len(words := segment["words"]) > 0:
+            saved_tokens += len(timing.tokens)
+            word_index += 1
+
+        if len(words) > 0:
             # adjust the segment-level timestamps based on the word-level timestamps
             segment["start"] = words[0]["start"]
             segment["end"] = words[-1]["end"]
+
+        segment["words"] = words

whisper/tokenizer.py

@@ -1,12 +1,12 @@
+import base64
 import os
 import string
-from dataclasses import dataclass
+from dataclasses import dataclass, field
 from functools import cached_property, lru_cache
-from typing import List, Optional, Tuple, Union
+from typing import Dict, List, Optional, Tuple
 
-import numpy as np
-import torch
-from transformers import GPT2TokenizerFast
+import tiktoken
+from tiktoken_ext.openai_public import gpt2
 
 LANGUAGES = {
     "en": "english",
@@ -127,74 +127,84 @@ TO_LANGUAGE_CODE = {
 }
 
 
-@dataclass(frozen=True)
+@dataclass
 class Tokenizer:
-    """A thin wrapper around `GPT2TokenizerFast` providing quick access to special tokens"""
+    """A thin wrapper around `tiktoken` providing quick access to special tokens"""
 
-    tokenizer: "GPT2TokenizerFast"
-    language: Optional[str]
-    sot_sequence: Tuple[int]
+    encoding: tiktoken.Encoding
+    language: Optional[str] = None
+    task: Optional[str] = None
+    sot_sequence: Tuple[int] = ()
+    special_tokens: Dict[str, int] = field(default_factory=dict)
+
+    def __post_init__(self):
+        for special in self.encoding.special_tokens_set:
+            special_token = self.encoding.encode_single_token(special)
+            self.special_tokens[special] = special_token
+
+        sot: int = self.special_tokens["<|startoftranscript|>"]
+        translate: int = self.special_tokens["<|translate|>"]
+        transcribe: int = self.special_tokens["<|transcribe|>"]
+
+        langs = tuple(LANGUAGES.keys())
+        sot_sequence = [sot]
+        if self.language is not None:
+            sot_sequence.append(sot + 1 + langs.index(self.language))
+        if self.task is not None:
+            task_token: int = transcribe if self.task == "transcribe" else translate
+            sot_sequence.append(task_token)
+
+        self.sot_sequence = tuple(sot_sequence)
 
     def encode(self, text, **kwargs):
-        return self.tokenizer.encode(text, **kwargs)
+        return self.encoding.encode(text, **kwargs)
 
-    def decode(
-        self, token_ids: Union[int, List[int], np.ndarray, torch.Tensor], **kwargs
-    ):
-        return self.tokenizer.decode(token_ids, **kwargs)
+    def decode(self, token_ids: List[int], **kwargs) -> str:
+        token_ids = [t for t in token_ids if t < self.timestamp_begin]
+        return self.encoding.decode(token_ids, **kwargs)
 
-    def decode_with_timestamps(self, tokens) -> str:
+    def decode_with_timestamps(self, token_ids: List[int], **kwargs) -> str:
         """
-        Timestamp tokens are above the special tokens' id range and are ignored by `decode()`.
+        Timestamp tokens are above other special tokens' id range and are ignored by `decode()`.
         This method decodes given tokens with timestamps tokens annotated, e.g. "<|1.08|>".
         """
-        outputs = [[]]
-        for token in tokens:
-            if token >= self.timestamp_begin:
-                timestamp = f"<|{(token - self.timestamp_begin) * 0.02:.2f}|>"
-                outputs.append(timestamp)
-                outputs.append([])
-            else:
-                outputs[-1].append(token)
-        return "".join(
-            [s if isinstance(s, str) else self.tokenizer.decode(s) for s in outputs]
-        )
+        return self.encoding.decode(token_ids, **kwargs)
 
     @cached_property
     def eot(self) -> int:
-        return self.tokenizer.eos_token_id
+        return self.encoding.eot_token
 
     @cached_property
     def transcribe(self) -> int:
-        return self._get_single_token_id("<|transcribe|>")
+        return self.special_tokens["<|transcribe|>"]
 
     @cached_property
     def translate(self) -> int:
-        return self._get_single_token_id("<|translate|>")
+        return self.special_tokens["<|translate|>"]
 
     @cached_property
     def sot(self) -> int:
-        return self._get_single_token_id("<|startoftranscript|>")
+        return self.special_tokens["<|startoftranscript|>"]
 
     @cached_property
     def sot_lm(self) -> int:
-        return self._get_single_token_id("<|startoflm|>")
+        return self.special_tokens["<|startoflm|>"]
 
     @cached_property
     def sot_prev(self) -> int:
-        return self._get_single_token_id("<|startofprev|>")
+        return self.special_tokens["<|startofprev|>"]
 
     @cached_property
     def no_speech(self) -> int:
-        return self._get_single_token_id("<|nospeech|>")
+        return self.special_tokens["<|nospeech|>"]
 
     @cached_property
     def no_timestamps(self) -> int:
-        return self._get_single_token_id("<|notimestamps|>")
+        return self.special_tokens["<|notimestamps|>"]
 
     @cached_property
     def timestamp_begin(self) -> int:
-        return self.tokenizer.all_special_ids[-1] + 1
+        return self.special_tokens["<|0.00|>"]
 
     @cached_property
     def language_token(self) -> int:
@@ -202,25 +212,15 @@ class Tokenizer:
         if self.language is None:
             raise ValueError("This tokenizer does not have language token configured")
 
-        additional_tokens = dict(
-            zip(
-                self.tokenizer.additional_special_tokens,
-                self.tokenizer.additional_special_tokens_ids,
-            )
-        )
-        candidate = f"<|{self.language}|>"
-        if candidate in additional_tokens:
-            return additional_tokens[candidate]
+        if token := self.special_tokens.get(f"<|{self.language}|>", None):
+            return token
 
         raise KeyError(f"Language {self.language} not found in tokenizer.")
 
     @cached_property
     def all_language_tokens(self) -> Tuple[int]:
         result = []
-        for token, token_id in zip(
-            self.tokenizer.additional_special_tokens,
-            self.tokenizer.additional_special_tokens_ids,
-        ):
+        for token, token_id in self.special_tokens.items():
             if token.strip("<|>") in LANGUAGES:
                 result.append(token_id)
         return tuple(result)
@@ -258,22 +258,17 @@ class Tokenizer:
         assert all(0x2640 <= ord(c) <= 0x267F for c in miscellaneous)
 
         # allow hyphens "-" and single quotes "'" between words, but not at the beginning of a word
-        result = {self.tokenizer.encode(" -")[0], self.tokenizer.encode(" '")[0]}
+        result = {self.encoding.encode(" -")[0], self.encoding.encode(" '")[0]}
         for symbol in symbols + list(miscellaneous):
             for tokens in [
-                self.tokenizer.encode(symbol),
-                self.tokenizer.encode(" " + symbol),
+                self.encoding.encode(symbol),
+                self.encoding.encode(" " + symbol),
             ]:
                 if len(tokens) == 1 or symbol in miscellaneous:
                     result.add(tokens[0])
 
         return tuple(sorted(result))
 
-    def _get_single_token_id(self, text) -> int:
-        tokens = self.tokenizer.encode(text)
-        assert len(tokens) == 1, f"{text} is not encoded as a single token"
-        return tokens[0]
-
     def split_to_word_tokens(self, tokens: List[int]):
         if self.language in {"zh", "ja", "th", "lo", "my"}:
             # These languages don't typically use spaces, so it is difficult to split words
@@ -284,17 +279,27 @@ class Tokenizer:
         return self.split_tokens_on_spaces(tokens)
 
     def split_tokens_on_unicode(self, tokens: List[int]):
+        decoded_full = self.decode_with_timestamps(tokens)
+        replacement_char = "\ufffd"
+
         words = []
         word_tokens = []
         current_tokens = []
+        unicode_offset = 0
 
         for token in tokens:
             current_tokens.append(token)
             decoded = self.decode_with_timestamps(current_tokens)
-            if "\ufffd" not in decoded:
+
+            if (
+                replacement_char not in decoded
+                or decoded_full[unicode_offset + decoded.index(replacement_char)]
+                == replacement_char
+            ):
                 words.append(decoded)
                 word_tokens.append(current_tokens)
                 current_tokens = []
+                unicode_offset += len(decoded)
 
         return words, word_tokens
 
@@ -318,12 +323,17 @@ class Tokenizer:
 
 
 @lru_cache(maxsize=None)
-def build_tokenizer(name: str = "gpt2"):
-    os.environ["TOKENIZERS_PARALLELISM"] = "false"
-    path = os.path.join(os.path.dirname(__file__), "assets", name)
-    tokenizer = GPT2TokenizerFast.from_pretrained(path)
+def get_encoding(name: str = "gpt2"):
+    vocab_path = os.path.join(os.path.dirname(__file__), "assets", f"{name}.tiktoken")
+    ranks = {
+        base64.b64decode(token): int(rank)
+        for token, rank in (line.split() for line in open(vocab_path) if line)
+    }
+    n_vocab = len(ranks)
+    special_tokens = {}
 
     specials = [
+        "<|endoftext|>",
         "<|startoftranscript|>",
         *[f"<|{lang}|>" for lang in LANGUAGES.keys()],
         "<|translate|>",
@@ -332,18 +342,28 @@ def build_tokenizer(name: str = "gpt2"):
         "<|startofprev|>",
         "<|nospeech|>",
         "<|notimestamps|>",
+        *[f"<|{i * 0.02:.2f}|>" for i in range(1501)],
     ]
 
-    tokenizer.add_special_tokens(dict(additional_special_tokens=specials))
-    return tokenizer
+    for token in specials:
+        special_tokens[token] = n_vocab
+        n_vocab += 1
+
+    return tiktoken.Encoding(
+        name=os.path.basename(vocab_path),
+        explicit_n_vocab=n_vocab,
+        pat_str=gpt2()["pat_str"],
+        mergeable_ranks=ranks,
+        special_tokens=special_tokens,
+    )
 
 
 @lru_cache(maxsize=None)
 def get_tokenizer(
     multilingual: bool,
     *,
-    task: Optional[str] = None,  # Literal["transcribe", "translate", None]
     language: Optional[str] = None,
+    task: Optional[str] = None,  # Literal["transcribe", "translate", None]
 ) -> Tokenizer:
     if language is not None:
         language = language.lower()
@@ -354,27 +374,14 @@ def get_tokenizer(
                 raise ValueError(f"Unsupported language: {language}")
 
     if multilingual:
-        tokenizer_name = "multilingual"
-        task = task or "transcribe"
+        encoding_name = "multilingual"
         language = language or "en"
+        task = task or "transcribe"
     else:
-        tokenizer_name = "gpt2"
-        task = None
+        encoding_name = "gpt2"
         language = None
+        task = None
 
-    tokenizer = build_tokenizer(name=tokenizer_name)
-    all_special_ids: List[int] = tokenizer.all_special_ids
-    sot: int = all_special_ids[1]
-    translate: int = all_special_ids[-6]
-    transcribe: int = all_special_ids[-5]
+    encoding = get_encoding(name=encoding_name)
 
-    langs = tuple(LANGUAGES.keys())
-    sot_sequence = [sot]
-    if language is not None:
-        sot_sequence.append(sot + 1 + langs.index(language))
-    if task is not None:
-        sot_sequence.append(transcribe if task == "transcribe" else translate)
-
-    return Tokenizer(
-        tokenizer=tokenizer, language=language, sot_sequence=tuple(sot_sequence)
-    )
+    return Tokenizer(encoding=encoding, language=language, task=task)

whisper/transcribe.py

@@ -200,14 +200,14 @@ def transcribe(
     def new_segment(
         *, start: float, end: float, tokens: torch.Tensor, result: DecodingResult
     ):
-        text_tokens = [token for token in tokens.tolist() if token < tokenizer.eot]
+        tokens = tokens.tolist()
+        text_tokens = [token for token in tokens if token < tokenizer.eot]
         return {
-            "id": len(all_segments),
             "seek": seek,
             "start": start,
             "end": end,
             "text": tokenizer.decode(text_tokens),
-            "tokens": text_tokens,
+            "tokens": tokens,
             "temperature": result.temperature,
             "avg_logprob": result.avg_logprob,
             "compression_ratio": result.compression_ratio,
@@ -245,7 +245,6 @@ def transcribe(
 
             previous_seek = seek
             current_segments = []
-            current_tokens = []
 
             timestamp_tokens: torch.Tensor = tokens.ge(tokenizer.timestamp_begin)
             single_timestamp_ending = timestamp_tokens[-2:].tolist() == [False, True]
@@ -275,7 +274,6 @@ def transcribe(
                             result=result,
                         )
                     )
-                    current_tokens.append(sliced_tokens.tolist())
                     last_slice = current_slice
 
                 if single_timestamp_ending:
@@ -287,7 +285,6 @@ def transcribe(
                         tokens[last_slice - 1].item() - tokenizer.timestamp_begin
                     )
                     seek += last_timestamp_pos * input_stride
-                all_tokens.extend(tokens[: last_slice + 1].tolist())
             else:
                 duration = segment_duration
                 timestamps = tokens[timestamp_tokens.nonzero().flatten()]
@@ -309,7 +306,6 @@ def transcribe(
                         result=result,
                     )
                 )
-                current_tokens.append(tokens.tolist())
                 seek += segment_size
 
             if not condition_on_previous_text or result.temperature > 0.5:
@@ -348,11 +344,17 @@ def transcribe(
                     segment["text"] = ""
                     segment["tokens"] = []
                     segment["words"] = []
-                    current_tokens[i] = []
 
-            all_segments.extend(current_segments)
+            all_segments.extend(
+                [
+                    {"id": i, **segment}
+                    for i, segment in enumerate(
+                        current_segments, start=len(all_segments)
+                    )
+                ]
+            )
             all_tokens.extend(
-                [token for segment in current_tokens for token in segment]
+                [token for segment in current_segments for token in segment["tokens"]]
             )
 
             # update progress bar

whisper/version.py

@@ -1 +1 @@
-__version__ = "20230307"
+__version__ = "20230314"