o ̳i=@sjddlmZmZmZmZmZddlmZmZddl m Z ddl m Z m Z mZgdZGddde e ZdS) )AnyListMappingOptionalTuple)MessagePromptTemplate) Transform)ModelTokenizerSentencePieceBaseTokenizer#tokenize_messages_no_special_tokens)     c @seZdZdZ  d$dedeedeefddZe dd Z e d d Z e d d Z e ddZ   d%dededededeef ddZdeedefddZdddeededeeeeeffddZ d&d eeefd!edeeeffd"d#ZdS)'GemmaTokenizera Gemma's implementation of the SentencePiece tokenizer Args: path (str): Path to pretrained tokenizer file. max_seq_len (Optional[int]): A max sequence length to truncate tokens to. Default: None prompt_template (Optional[PromptTemplate]): template used to format the messages based on their role. This is used to add structured text around the actual messages. The structured text is used in three scenarios: - Task-specific templates to gear models for a particular task that it will expect after training - Model-specific templates that are required whenever the model is prompted, such as the [INST] tags in Llama2 and in Mistral - Community standardized templates, such as :class:`~torchtune.data.ChatMLTemplate` The extra text will still get tokenized as normal text, not as special tokens. Default is None. Examples: >>> tokenizer = GemmaTokenizer("/path/to/spm_model") >>> tokenized_text = tokenizer.encode("Hello world!", add_bos=True, add_eos=True) >>> print(tokenized_text) [1, 31587, 29644, 102, 2] Npath max_seq_lenprompt_templatecCs,t||_d|j_|jg|_||_||_dS)Nr)r _spm_modelpad_ideos_id stop_tokensrr)selfrrrrU/home/ubuntu/.local/lib/python3.10/site-packages/torchtune/models/gemma/_tokenizer.py__init__-s   zGemmaTokenizer.__init__cC|jjSN)rrrrrrr?zGemmaTokenizer.eos_idcCrr)rbos_idr rrrr"Cr!zGemmaTokenizer.bos_idcCrr)rrr rrrrGr!zGemmaTokenizer.pad_idcCrr)r vocab_sizer rrrr#Kr!zGemmaTokenizer.vocab_sizeTFtextadd_bosadd_eostrim_leading_whitespacereturncCs|jj||||dS)N)r%r&r')rencode)rr$r%r&r'rrrr)Os zGemmaTokenizer.encode token_idscCs |j|Sr)rdecode)rr*rrrr+]s zGemmaTokenizer.decode)r&messagescCs8|jdur ||n|}t|||j|r|jdSddS)aTokenize a list of messages one at a time then concatenate them, returning a list of tokens and a list of masks. Example: >>> tokenizer = GemmaTokenizer(tokenizer_path, max_seq_len) >>> messages = [ Message(role="system", content="system message\n", masked=True), Message(role="user", content="user prompt\n", masked=True), Message(role="assistant", content="assistant response\n"), ] >>> # tokenize_messages encodes messages separately and concats >>> tokenizer.tokenize_messages(messages)[0] [1, 1788, 2643, 13, 1792, 9508, 13, 465, 22137, 2933, 2] >>> # Same result as encoding the full string in one go >>> tokenizer.encode(''.join([message.content for message in messages])) [1, 1788, 2643, 13, 1792, 9508, 13, 465, 22137, 2933, 2] Args: messages (List[Message]): A list of messages, each containing role, content, and masked attributes. add_eos (bool): Whether to append EOS after assistant message, default to True Returns: Tuple[List[int], List[bool]]: The tokenized messages N) tokenizerr,r"r)rr r"r)rr,r&templated_messagesrrrtokenize_messagescs & z GemmaTokenizer.tokenize_messagessample inferencecCs,|d}||\}}||d<||d<|S)a Apply ``tokenize_messages`` to the "messages" field in the sample. Args: sample (Mapping[str, Any]): A sample with a "messages" field containing a List[Message] to tokenize inference (bool): Whether the template is being used for inference or not. Returns: Mapping[str, Any]: The sample with added "tokens" and "mask" fields and the "messages" field removed. r,tokensmask)popr/)rr0r1r,r2r3rrr__call__s zGemmaTokenizer.__call__)NN)TTF)F)__name__ __module__ __qualname____doc__strrintrrpropertyrr"rr#boolrr)r+rrr/rrr5rrrrrsl         1  rN)typingrrrrrtorchtune.datarrtorchtune.modules.transformsr 'torchtune.modules.transforms.tokenizersr r r WHITESPACE_CHARSrrrrrs