o Ti+@sDddlmZmZddlmZmZddlmZeGdddeZdS)) dataclassfield)AnyOptional)TrainingArgumentscseZdZUdZejddgZedddidZee d<ed dd idZ ee d <ed dd idZ e e e d<edddidZe ee d<edddidZe ee d<ed ddidZe ee d<edddidZee d<eddddgddZee d<eddd idZee d!<eddd"idZee d#<ed$dd%idZee d&<ed dd'idZe ee d(<ed)d*d)d+gddZee d,<ed-dd.idZe e d/<ed dd0idZe e e d1<ed2dd3idZe e d4<ed-dd5idZe e d6<ed dd7idZe eee fe d<ed dd8idZ!e eee fe d<ed dd9idZ"e ee d:<ed-dd;idZ#e e d<<ed=dd>idZ$ee d?<fd@dAZ%Z&S)B KTOConfigu Configuration class for the [`KTOTrainer`]. This class includes only the parameters that are specific to KTO training. For a full list of training arguments, please refer to the [`~transformers.TrainingArguments`] documentation. Note that default values in this class may differ from those in [`~transformers.TrainingArguments`]. Using [`~transformers.HfArgumentParser`] we can turn this class into [argparse](https://docs.python.org/3/library/argparse#module-argparse) arguments that can be specified on the command line. Parameters: max_length (`int` or `None`, *optional*, defaults to `1024`): Maximum length of the sequences (prompt + completion) in the batch. This argument is required if you want to use the default data collator. max_prompt_length (`int` or `None`, *optional*, defaults to `512`): Maximum length of the prompt. This argument is required if you want to use the default data collator. max_completion_length (`int` or `None`, *optional*, defaults to `None`): Maximum length of the completion. This argument is required if you want to use the default data collator and your model is an encoder-decoder. beta (`float`, *optional*, defaults to `0.1`): Parameter controlling the deviation from the reference model. Higher β means less deviation from the reference model. loss_type (`str`, *optional*, defaults to `"kto"`): Type of loss to use. Possible values are: - `"kto"`: KTO loss from the [KTO](https://huggingface.co/papers/2402.01306) paper. - `"apo_zero_unpaired"`: Unpaired variant of APO-zero loss from the [APO](https://huggingface.co/papers/2408.06266) paper. desirable_weight (`float`, *optional*, defaults to `1.0`): Desirable losses are weighed by this factor to counter unequal number of desirable and undesirable paris. undesirable_weight (`float`, *optional*, defaults to `1.0`): Undesirable losses are weighed by this factor to counter unequal number of desirable and undesirable pairs. label_pad_token_id (`int`, *optional*, defaults to `-100`): Label pad token id. This argument is required if you want to use the default data collator. padding_value (`int` or `None`, *optional*, defaults to `None`): Padding value to use. If `None`, the padding value of the tokenizer is used. truncation_mode (`str`, *optional*, defaults to `"keep_end"`): Truncation mode to use when the prompt is too long. Possible values are `"keep_end"` or `"keep_start"`. This argument is required if you want to use the default data collator. generate_during_eval (`bool`, *optional*, defaults to `False`): If `True`, generates and logs completions from both the model and the reference model to W&B or Comet during evaluation. is_encoder_decoder (`bool` or `None`, *optional*, defaults to `None`): When using the `model_init` argument (callable) to instantiate the model instead of the `model` argument, you need to specify if the model returned by the callable is an encoder-decoder model. precompute_ref_log_probs (`bool`, *optional*, defaults to `False`): Whether to precompute reference model log probabilities for training and evaluation datasets. This is useful when training without the reference model to reduce the total GPU memory needed. model_init_kwargs (`dict[str, Any]` or `None`, *optional*, defaults to `None`): Keyword arguments to pass to `AutoModelForCausalLM.from_pretrained` when instantiating the model from a string. ref_model_init_kwargs (`dict[str, Any]` or `None`, *optional*, defaults to `None`): Keyword arguments to pass to `AutoModelForCausalLM.from_pretrained` when instantiating the reference model from a string. dataset_num_proc: (`int` or `None`, *optional*, defaults to `None`): Number of processes to use for processing the dataset. disable_dropout (`bool`, *optional*, defaults to `True`): Whether to disable dropout in the model and reference model. use_liger_loss (`bool`, *optional*, defaults to `False`): Whether to use Liger loss. It requires liger-kernel to be installed. base_model_attribute_name (`str`, *optional*, defaults to `"model"`): Name of the attribute in the model that contains the base model. This is used to get the base model from the model when the model does not have a `get_decoder` method in the case when `use_liger_loss` is `True`. model_init_kwargsref_model_init_kwargsgư>helpz$The initial learning rate for AdamW.)defaultmetadata learning_rate zLog every X updates steps. Should be an integer or a float in range `[0,1)`. If smaller than 1, will be interpreted as ratio of total training steps. logging_stepsNzWhether to use bf16 (mixed) precision instead of 32-bit. Requires Ampere or higher NVIDIA architecture or Intel XPU or using CPU (use_cpu) or Ascend NPU. If not set, it defaults to `True` if `fp16` is not set.bf16izCMaximum length of the sequences (prompt + completion) in the batch. max_lengthizMaximum length of the prompt. This argument is required if you want to use the default data collator and your model is an encoder-decoder.max_prompt_lengthzMaximum length of the completion. This argument is required if you want to use the default data collator and your model is an encoder-decoder.max_completion_lengthg?uvParameter controlling the deviation from the reference model. Higher β means less deviation from the reference model.betaktozType of loss to use.apo_zero_unpaired)r choices loss_typeg?ziDesirable losses are weighed by this factor to counter unequal number of desirable and undesirable pairs.desirable_weightzkUndesirable losses are weighed by this factor to counter unequal number of desirable and undesirable pairs.undesirable_weightiz[Label pad token id. This argument is required if you want to use the default data collator.label_pad_token_idzLPadding value to use. If `None`, the padding value of the tokenizer is used. padding_valuekeep_endz3Truncation mode to use when the prompt is too long. keep_starttruncation_modeFzoIf `True`, generates and logs completions from both the model and the reference model to W&B during evaluation.generate_during_evalzWhen using the `model_init` argument (callable) to instantiate the model instead of the `model` argument, you need to specify if the model returned by the callable is an encoder-decoder model.is_encoder_decoderTz(Whether to disable dropout in the model.disable_dropoutzWhether to precompute reference model log probabilities for training and evaluation datasets. This is useful when training without the reference model to reduce the total GPU memory needed.precompute_ref_log_probszoKeyword arguments to pass to `AutoModelForCausalLM.from_pretrained` when instantiating the model from a string.zyKeyword arguments to pass to `AutoModelForCausalLM.from_pretrained` when instantiating the reference model from a string.z6Number of processes to use for processing the dataset.dataset_num_proczDWhether to use Liger loss. It requires liger-kernel to be installed.use_liger_lossmodelzName of the attribute in the model that contains the base model. This is used to get the base model from the model when the model does not have a `get_decoder` method in the case when `use_liger_loss` is `True`.base_model_attribute_namecs(|jdur |j n|j|_tdS)N)rfp16super __post_init__)self __class__J/home/ubuntu/.local/lib/python3.10/site-packages/trl/trainer/kto_config.pyr*szKTOConfig.__post_init__)'__name__ __module__ __qualname____doc__r_VALID_DICT_FIELDSrr float__annotations__rrrboolrintrrrrstrrrrrrr r!r"r#rdictrr r$r%r'r* __classcell__r.r.r,r/rs C  rN) dataclassesrrtypingrr transformersrrr.r.r.r/s