From 2840f0a51e3fc509b049a0cb82f0a5f75a38c2b1 Mon Sep 17 00:00:00 2001 From: "renxuancheng.rxc" Date: Wed, 17 Jan 2024 11:53:35 +0800 Subject: [PATCH 1/2] add warning for sliding window attention --- src/transformers/models/qwen2/modeling_qwen2.py | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/src/transformers/models/qwen2/modeling_qwen2.py b/src/transformers/models/qwen2/modeling_qwen2.py index f1c31e28a446..aced4c2add8c 100644 --- a/src/transformers/models/qwen2/modeling_qwen2.py +++ b/src/transformers/models/qwen2/modeling_qwen2.py @@ -729,6 +729,11 @@ def __init__(self, config: Qwen2Config, layer_idx: int): super().__init__() self.hidden_size = config.hidden_size + if config.use_sliding_window and config._attn_implementation != "flash_attention_2": + logger.warning_once( + f"Sliding Window Attention is enabled but not implemented for `{config._attn_implementation}`; " + "unexpected results may be encountered." + ) self.self_attn = QWEN2_ATTENTION_CLASSES[config._attn_implementation](config, layer_idx) self.mlp = Qwen2MLP(config) From 87f6cf70e681f4704cbce339c84a42a9c804026d Mon Sep 17 00:00:00 2001 From: "renxuancheng.rxc" Date: Wed, 17 Jan 2024 11:54:06 +0800 Subject: [PATCH 2/2] update qwen2 tokenization --- src/transformers/convert_slow_tokenizer.py | 2 +- .../models/qwen2/tokenization_qwen2_fast.py | 16 +++++++++++++++- tests/models/qwen2/test_tokenization_qwen2.py | 15 +-------------- 3 files changed, 17 insertions(+), 16 deletions(-) diff --git a/src/transformers/convert_slow_tokenizer.py b/src/transformers/convert_slow_tokenizer.py index 7ec3cc8e00d6..46f2b2dc2309 100644 --- a/src/transformers/convert_slow_tokenizer.py +++ b/src/transformers/convert_slow_tokenizer.py @@ -385,7 +385,7 @@ def converted(self) -> Tokenizer: invert=False, ), pre_tokenizers.ByteLevel( - add_prefix_space=False, # The slow tokenizer does not support add_prefix_space + add_prefix_space=getattr(self.original_tokenizer, "add_prefix_space", False), use_regex=False, ), ] diff --git a/src/transformers/models/qwen2/tokenization_qwen2_fast.py b/src/transformers/models/qwen2/tokenization_qwen2_fast.py index 6830d9c9dd86..e5f20f210352 100644 --- a/src/transformers/models/qwen2/tokenization_qwen2_fast.py +++ b/src/transformers/models/qwen2/tokenization_qwen2_fast.py @@ -71,6 +71,13 @@ class Qwen2TokenizerFast(PreTrainedTokenizerFast): tokenizer_file (`str`, *optional*): Path to [tokenizers](https://github.com/huggingface/tokenizers) file (generally has a .json extension) that contains everything needed to load the tokenizer. + unk_token (`str`, *optional*): + The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this + token instead. Not applicable to this tokenizer. + bos_token (`str`, *optional*): + The beginning of sequence token. Not applicable for this tokenizer. + eos_token (`str`, *optional*, defaults to `"<|endoftext|>"`): + The end of sequence token. """ vocab_files_names = VOCAB_FILES_NAMES @@ -84,15 +91,22 @@ def __init__( vocab_file=None, merges_file=None, tokenizer_file=None, + unk_token=None, + bos_token=None, + eos_token="<|endoftext|>", **kwargs, ): # We need to at least pass vocab_file and merges_file to base class # in case a slow tokenizer needs to be initialized; other can be - # configured through files + # configured through files. + # following GPT2TokenizerFast, also adding unk_token, bos_token, and eos_token super().__init__( vocab_file, merges_file, tokenizer_file=tokenizer_file, + unk_token=unk_token, + bos_token=bos_token, + eos_token=eos_token, **kwargs, ) diff --git a/tests/models/qwen2/test_tokenization_qwen2.py b/tests/models/qwen2/test_tokenization_qwen2.py index 6a6469caa9ad..fd40556e03eb 100644 --- a/tests/models/qwen2/test_tokenization_qwen2.py +++ b/tests/models/qwen2/test_tokenization_qwen2.py @@ -194,20 +194,7 @@ def test_tokenizer_integration(self): """tokenizer("世界，你好！")```""", ] - # fmt: off - expected_encoding = { - 'input_ids': [ - [8963, 388, 320, 69514, 3881, 438, 4510, 27414, 32852, 388, 323, 4510, 27414, 21334, 35722, 1455, 529, 8, 5707, 4586, 58238, 77235, 320, 61437, 11, 479, 2828, 12, 17, 11, 11830, 61437, 64, 11, 1599, 10994, 11, 27604, 321, 33, 529, 11, 29881, 6954, 32574, 369, 18448, 11434, 45451, 320, 45, 23236, 8, 323, 18448, 11434, 23470, 320, 30042, 38, 8, 448, 916, 220, 18, 17, 10, 80669, 4119, 304, 220, 16, 15, 15, 10, 15459, 323, 5538, 94130, 2897, 1948, 619, 706, 11, 5355, 51, 21584, 323, 94986, 13], - [144834, 80532, 93685, 83744, 34187, 73670, 104261, 29490, 62189, 103937, 104034, 102830, 98841, 104034, 104949, 9370, 5333, 58143, 102011, 1773, 37029, 98841, 104034, 104949, 73670, 101940, 100768, 104997, 33108, 100912, 105054, 90395, 100136, 106831, 45181, 64355, 104034, 113521, 101975, 33108, 85329, 1773, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643], - [73594, 12669, 198, 85593, 284, 8979, 37434, 6387, 10442, 35722, 445, 48, 16948, 45274, 16948, 34841, 3135, 1138, 85593, 445, 99489, 3837, 108386, 6313, 899, 73594, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643], - ], - 'attention_mask': [ - [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], - [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], - [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], - ], - } - # fmt: on + expected_encoding = {'input_ids': [[8963, 388, 320, 69514, 3881, 438, 4510, 27414, 32852, 388, 323, 4510, 27414, 21334, 35722, 1455, 529, 8, 5707, 4586, 58238, 77235, 320, 61437, 11, 479, 2828, 12, 17, 11, 11830, 61437, 64, 11, 1599, 10994, 11, 27604, 321, 33, 529, 11, 29881, 6954, 32574, 369, 18448, 11434, 45451, 320, 45, 23236, 8, 323, 18448, 11434, 23470, 320, 30042, 38, 8, 448, 916, 220, 18, 17, 10, 80669, 4119, 304, 220, 16, 15, 15, 10, 15459, 323, 5538, 94130, 2897, 1948, 619, 706, 11, 5355, 51, 21584, 323, 94986, 13], [144834, 80532, 93685, 83744, 34187, 73670, 104261, 29490, 62189, 103937, 104034, 102830, 98841, 104034, 104949, 9370, 5333, 58143, 102011, 1773, 37029, 98841, 104034, 104949, 73670, 101940, 100768, 104997, 33108, 100912, 105054, 90395, 100136, 106831, 45181, 64355, 104034, 113521, 101975, 33108, 85329, 1773, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643], [73594, 12669, 198, 85593, 284, 8979, 37434, 6387, 10442, 35722, 445, 48, 16948, 45274, 16948, 34841, 3135, 1138, 85593, 445, 99489, 3837, 108386, 6313, 899, 73594, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643, 151643]], 'attention_mask': [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # fmt: off self.tokenizer_integration_test_util( expected_encoding=expected_encoding,