The Trainer class provides an API for feature-complete training in PyTorch, and it supports distributed training on multiple GPUs/TPUs, mixed precision for NVIDIA GPUs, AMD GPUs, and torch.amp for PyTorch. Trainer goes hand-in-hand with the TrainingArguments class, which offers a wide range of options to customize how a model is trained. Together, these two classes provide a complete training API.
Seq2SeqTrainer and Seq2SeqTrainingArguments inherit from the Trainer and TrainingArgument classes and they’re adapted for training models for sequence-to-sequence tasks such as summarization or translation.
The Trainer class is optimized for 🤗 Transformers models and can have surprising behaviors when used with other models. When using it with your own model, make sure:
labels argument is provided and that loss is returned as the first
element of the tuple (if your model returns tuples)label_names in TrainingArguments to indicate their name to the Trainer) but none of them should be named "label"( model: Union = None args: TrainingArguments = None data_collator: Optional = None train_dataset: Optional = None eval_dataset: Union = None tokenizer: Optional = None model_init: Optional = None compute_metrics: Optional = None callbacks: Optional = None optimizers: Tuple = (None, None) preprocess_logits_for_metrics: Optional = None )
Parameters
torch.nn.Module, optional) —
The model to train, evaluate or use for predictions. If not provided, a model_init must be passed.
Trainer is optimized to work with the PreTrainedModel provided by the library. You can still use
your own models defined as torch.nn.Module as long as they work the same way as the 🤗 Transformers
models.
output_dir set to a directory named tmp_trainer in the current directory if not provided. DataCollator, optional) —
The function to use to form a batch from a list of elements of train_dataset or eval_dataset. Will
default to default_data_collator() if no tokenizer is provided, an instance of
DataCollatorWithPadding otherwise. torch.utils.data.Dataset or torch.utils.data.IterableDataset, optional) —
The dataset to use for training. If it is a Dataset, columns not accepted by the
model.forward() method are automatically removed.
Note that if it’s a torch.utils.data.IterableDataset with some randomization and you are training in a
distributed fashion, your iterable dataset should either use a internal attribute generator that is a
torch.Generator for the randomization that must be identical on all processes (and the Trainer will
manually set the seed of this generator at each epoch) or have a set_epoch() method that internally
sets the seed of the RNGs used.
torch.utils.data.Dataset, Dict[str, torch.utils.data.Dataset]), optional) —
The dataset to use for evaluation. If it is a Dataset, columns not accepted by the
model.forward() method are automatically removed. If it is a dictionary, it will evaluate on each
dataset prepending the dictionary key to the metric name. Callable[[], PreTrainedModel], optional) —
A function that instantiates the model to be used. If provided, each call to train() will start
from a new instance of the model as given by this function.
The function may have zero argument, or a single one containing the optuna/Ray Tune/SigOpt trial object, to be able to choose different architectures according to hyper parameters (such as layer count, sizes of inner layers, dropout probabilities etc).
Callable[[EvalPrediction], Dict], optional) —
The function that will be used to compute metrics at evaluation. Must take a EvalPrediction and return
a dictionary string to metric values. If you want to remove one of the default callbacks used, use the Trainer.remove_callback() method.
Tuple[torch.optim.Optimizer, torch.optim.lr_scheduler.LambdaLR], optional, defaults to (None, None)) —
A tuple containing the optimizer and the scheduler to use. Will default to an instance of AdamW on your
model and a scheduler given by get_linear_schedule_with_warmup() controlled by args. Callable[[torch.Tensor, torch.Tensor], torch.Tensor], optional) —
A function that preprocess the logits right before caching them at each evaluation step. Must take two
tensors, the logits and the labels, and return the logits once processed as desired. The modifications made
by this function will be reflected in the predictions received by compute_metrics.
Note that the labels (second parameter) will be None if the dataset does not have them.
Trainer is a simple but feature-complete training and eval loop for PyTorch, optimized for 🤗 Transformers.
Important attributes:
DeepSpeed,
the inner model is wrapped in DeepSpeed and then again in torch.nn.DistributedDataParallel. If the inner
model hasn’t been wrapped, then self.model_wrapped is the same as self.model.False if model parallel or deepspeed is used, or if the default
TrainingArguments.place_model_on_device is overridden to return False .train (e.g. when evaluate is called while
in train)( callback )
Parameters
type or TrainerCallback) —
A TrainerCallback class or an instance of a TrainerCallback. In the
first case, will instantiate a member of that class. Add a callback to the current list of TrainerCallback.
A helper wrapper that creates an appropriate context manager for autocast while feeding it the desired
arguments, depending on the situation.
How the loss is computed by Trainer. By default, all models return the loss in the first element.
Subclass and override for custom behavior.
A helper wrapper to group together context managers.
( language: Optional = None license: Optional = None tags: Union = None model_name: Optional = None finetuned_from: Optional = None tasks: Union = None dataset_tags: Union = None dataset: Union = None dataset_args: Union = None )
Parameters
str, optional) —
The language of the model (if applicable) str, optional) —
The license of the model. Will default to the license of the pretrained model used, if the original
model given to the Trainer comes from a repo on the Hub. str or List[str], optional) —
Some tags to be included in the metadata of the model card. str, optional) —
The name of the model. str, optional) —
The name of the model used to fine-tune this one (if applicable). Will default to the name of the repo
of the original model given to the Trainer (if it comes from the Hub). str or List[str], optional) —
One or several task identifiers, to be included in the metadata of the model card. str or List[str], optional) —
One or several dataset tags, to be included in the metadata of the model card. str or List[str], optional) —
One or several dataset identifiers, to be included in the metadata of the model card. str or List[str], optional) —
One or several dataset arguments, to be included in the metadata of the model card. Creates a draft of a model card using the information available to the Trainer.
Setup the optimizer.
We provide a reasonable default that works well. If you want to use something else, you can pass a tuple in the
Trainer’s init through optimizers, or subclass and override this method in a subclass.
Setup the optimizer and the learning rate scheduler.
We provide a reasonable default that works well. If you want to use something else, you can pass a tuple in the
Trainer’s init through optimizers, or subclass and override this method (or create_optimizer and/or
create_scheduler) in a subclass.
( num_training_steps: int optimizer: Optimizer = None )
Setup the scheduler. The optimizer of the trainer must have been set up either before this method is called or passed as an argument.
( eval_dataset: Union = None ignore_keys: Optional = None metric_key_prefix: str = 'eval' )
Parameters
Dataset, Dict[str, Dataset]), optional) —
Pass a dataset if you wish to override self.eval_dataset. If it is a Dataset, columns
not accepted by the model.forward() method are automatically removed. If it is a dictionary, it will
evaluate on each dataset, prepending the dictionary key to the metric name. Datasets must implement the
__len__ method.
If you pass a dictionary with names of datasets as keys and datasets as values, evaluate will run
separate evaluations on each dataset. This can be useful to monitor how training affects other
datasets or simply to get a more fine-grained evaluation.
When used with load_best_model_at_end, make sure metric_for_best_model references exactly one
of the datasets. If you, for example, pass in {"data1": data1, "data2": data2} for two datasets
data1 and data2, you could specify metric_for_best_model="eval_data1_loss" for using the
loss on data1 and metric_for_best_model="eval_data1_loss" for the loss on data2.
List[str], optional) —
A list of keys in the output of your model (if it is a dictionary) that should be ignored when
gathering predictions. str, optional, defaults to "eval") —
An optional prefix to be used as the metrics key prefix. For example the metrics “bleu” will be named
“eval_bleu” if the prefix is “eval” (default) Run evaluation and returns metrics.
The calling script will be responsible for providing a method to compute metrics, as they are task-dependent
(pass it to the init compute_metrics argument).
You can also subclass and override this method to inject custom behavior.
( dataloader: DataLoader description: str prediction_loss_only: Optional = None ignore_keys: Optional = None metric_key_prefix: str = 'eval' )
Prediction/evaluation loop, shared by Trainer.evaluate() and Trainer.predict().
Works both with or without labels.
( inputs: Dict ) → int
For models that inherit from PreTrainedModel, uses that method to compute the number of floating point operations for every backward + forward pass. If using another model, either implement such a method in the model or subclass and override this method.
Get all parameter names that weight decay will be applied to
Note that some models implement their own layernorm instead of calling nn.LayerNorm, weight decay could still apply to those modules since this function only filter out instance of nn.LayerNorm
( eval_dataset: Optional = None )
Parameters
torch.utils.data.Dataset, optional) —
If provided, will override self.eval_dataset. If it is a Dataset, columns not accepted
by the model.forward() method are automatically removed. It must implement __len__. Returns the evaluation ~torch.utils.data.DataLoader.
Subclass and override this method if you want to inject some custom behavior.
( args: TrainingArguments )
Returns the optimizer class and optimizer parameters based on the training arguments.
( test_dataset: Dataset )
Parameters
torch.utils.data.Dataset, optional) —
The test dataset to use. If it is a Dataset, columns not accepted by the
model.forward() method are automatically removed. It must implement __len__. Returns the test ~torch.utils.data.DataLoader.
Subclass and override this method if you want to inject some custom behavior.
Returns the training ~torch.utils.data.DataLoader.
Will use no sampler if train_dataset does not implement __len__, a random sampler (adapted to distributed
training if necessary) otherwise.
Subclass and override this method if you want to inject some custom behavior.
( hp_space: Optional = None compute_objective: Optional = None n_trials: int = 20 direction: Union = 'minimize' backend: Union = None hp_name: Optional = None **kwargs ) → [trainer_utils.BestRun or List[trainer_utils.BestRun]]
Parameters
Callable[["optuna.Trial"], Dict[str, float]], optional) —
A function that defines the hyperparameter search space. Will default to
default_hp_space_optuna() or default_hp_space_ray() or
default_hp_space_sigopt() depending on your backend. Callable[[Dict[str, float]], float], optional) —
A function computing the objective to minimize or maximize from the metrics returned by the evaluate
method. Will default to default_compute_objective(). int, optional, defaults to 100) —
The number of trial runs to test. str or List[str], optional, defaults to "minimize") —
If it’s single objective optimization, direction is str, can be "minimize" or "maximize", you
should pick "minimize" when optimizing the validation loss, "maximize" when optimizing one or
several metrics. If it’s multi objectives optimization, direction is List[str], can be List of
"minimize" and "maximize", you should pick "minimize" when optimizing the validation loss,
"maximize" when optimizing one or several metrics. str or ~training_utils.HPSearchBackend, optional) —
The backend to use for hyperparameter search. Will default to optuna or Ray Tune or SigOpt, depending
on which one is installed. If all are installed, will default to optuna. Callable[["optuna.Trial"], str]], optional) —
A function that defines the trial/run name. Will default to None. Dict[str, Any], optional) —
Additional keyword arguments passed along to optuna.create_study or ray.tune.run. For more
information see:
Returns
[trainer_utils.BestRun or List[trainer_utils.BestRun]]
All the information about the best run or best
runs for multi-objective optimization. Experiment summary can be found in run_summary attribute for Ray
backend.
Launch an hyperparameter search using optuna or Ray Tune or SigOpt. The optimized quantity is determined
by compute_objective, which defaults to a function returning the evaluation loss when no metric is provided,
the sum of all metrics otherwise.
To use this method, you need to have provided a model_init when initializing your Trainer: we need to
reinitialize the model at each new run. This is incompatible with the optimizers argument, so you need to
subclass Trainer and override the method create_optimizer_and_scheduler() for custom
optimizer/scheduler.
Initializes a git repo in self.args.hub_model_id.
Whether or not this process is the local (e.g., on one machine if training in a distributed fashion on several machines) main process.
Whether or not this process is the global main process (when training in a distributed fashion on several
machines, this is only going to be True for one process).
Log logs on the various objects watching training.
Subclass and override this method to inject custom behavior.
( split metrics )
Log metrics in a specially formatted way
Under distributed environment this is done only for a process with rank 0.
Notes on memory reports:
In order to get memory usage report you need to install psutil. You can do that with pip install psutil.
Now when this method is run, you will see a report that will include: :
init_mem_cpu_alloc_delta = 1301MB
init_mem_cpu_peaked_delta = 154MB
init_mem_gpu_alloc_delta = 230MB
init_mem_gpu_peaked_delta = 0MB
train_mem_cpu_alloc_delta = 1345MB
train_mem_cpu_peaked_delta = 0MB
train_mem_gpu_alloc_delta = 693MB
train_mem_gpu_peaked_delta = 7MBUnderstanding the reports:
train__, tells you which stage the metrics are for. Reports starting with init_
will be added to the first stage that gets run. So that if only evaluation is run, the memory usage for the
__init__ will be reported along with the eval_ metrics.cpu or gpu, tells you whether it’s the general RAM or the gpu0 memory
metric.*_alloc_delta - is the difference in the used/allocated memory counter between the end and the start of the
stage - it can be negative if a function released more memory than it allocated.*_peaked_delta - is any extra memory that was consumed and then freed - relative to the current allocated
memory counter - it is never negative. When you look at the metrics of any stage you add up alloc_delta +
peaked_delta and you know how much memory was needed to complete that stage.The reporting happens only for process of rank 0 and gpu 0 (if there is a gpu). Typically this is enough since the main process does the bulk of work, but it could be not quite so if model parallel is used and then other GPUs may use a different amount of gpu memory. This is also not the same under DataParallel where gpu0 may require much more memory than the rest since it stores the gradient and optimizer states for all participating GPUS. Perhaps in the future these reports will evolve to measure those too.
The CPU RAM metric measures RSS (Resident Set Size) includes both the memory which is unique to the process and the memory shared with other processes. It is important to note that it does not include swapped out memory, so the reports could be imprecise.
The CPU peak memory is measured using a sampling thread. Due to python’s GIL it may miss some of the peak memory if
that thread didn’t get a chance to run when the highest memory was used. Therefore this report can be less than
reality. Using tracemalloc would have reported the exact peak memory, but it doesn’t report memory allocations
outside of python. So if some C++ CUDA extension allocated its own memory it won’t be reported. And therefore it
was dropped in favor of the memory sampling approach, which reads the current process memory usage.
The GPU allocated and peak memory reporting is done with torch.cuda.memory_allocated() and
torch.cuda.max_memory_allocated(). This metric reports only “deltas” for pytorch-specific allocations, as
torch.cuda memory management system doesn’t track any memory allocated outside of pytorch. For example, the very
first cuda call typically loads CUDA kernels, which may take from 0.5 to 2GB of GPU memory.
Note that this tracker doesn’t account for memory allocations outside of Trainer’s __init__, train,
evaluate and predict calls.
Because evaluation calls may happen during train, we can’t handle nested invocations because
torch.cuda.max_memory_allocated is a single counter, so if it gets reset by a nested eval call, train’s tracker
will report incorrect info. If this pytorch issue gets resolved
it will be possible to change this class to be re-entrant. Until then we will only track the outer level of
train, evaluate and predict methods. Which means that if eval is called during train, it’s the latter
that will account for its memory usage and that of the former.
This also means that if any other tool that is used along the Trainer calls
torch.cuda.reset_peak_memory_stats, the gpu peak memory stats could be invalid. And the Trainer will disrupt
the normal behavior of any such tools that rely on calling torch.cuda.reset_peak_memory_stats themselves.
For best performance you may want to consider turning the memory profiling off for production runs.
( metrics: Dict ) → metrics (Dict[str, float])
Reformat Trainer metrics values to a human-readable format
Helper to get number of samples in a ~torch.utils.data.DataLoader by accessing its dataset. When
dataloader.dataset does not exist or has no length, estimates as best it can
Helper to get number of tokens in a ~torch.utils.data.DataLoader by enumerating dataloader.
( callback ) → TrainerCallback
Parameters
type or TrainerCallback) —
A TrainerCallback class or an instance of a TrainerCallback. In the
first case, will pop the first member of that class found in the list of callbacks. Returns
The callback removed, if found.
Remove a callback from the current list of TrainerCallback and returns it.
If the callback is not found, returns None (and no error is raised).
( test_dataset: Dataset ignore_keys: Optional = None metric_key_prefix: str = 'test' )
Parameters
Dataset) —
Dataset to run the predictions on. If it is an datasets.Dataset, columns not accepted by the
model.forward() method are automatically removed. Has to implement the method __len__ List[str], optional) —
A list of keys in the output of your model (if it is a dictionary) that should be ignored when
gathering predictions. str, optional, defaults to "test") —
An optional prefix to be used as the metrics key prefix. For example the metrics “bleu” will be named
“test_bleu” if the prefix is “test” (default) Run prediction and returns predictions and potential metrics.
Depending on the dataset and your use case, your test dataset may contain labels. In that case, this method
will also return metrics, like in evaluate().
If your predictions or labels have different sequence length (for instance because you’re doing dynamic padding in a token classification task) the predictions will be padded (on the right) to allow for concatenation into one array. The padding index is -100.
Returns: NamedTuple A namedtuple with the following keys:
np.ndarray): The predictions on test_dataset.np.ndarray, optional): The labels (if the dataset contained some).Dict[str, float], optional): The potential dictionary of metrics (if the dataset contained
labels).( dataloader: DataLoader description: str prediction_loss_only: Optional = None ignore_keys: Optional = None metric_key_prefix: str = 'eval' )
Prediction/evaluation loop, shared by Trainer.evaluate() and Trainer.predict().
Works both with or without labels.
( model: Module inputs: Dict prediction_loss_only: bool ignore_keys: Optional = None ) → Tuple[Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor]]
Parameters
nn.Module) —
The model to evaluate. Dict[str, Union[torch.Tensor, Any]]) —
The inputs and targets of the model.
The dictionary will be unpacked before being fed to the model. Most models expect the targets under the
argument labels. Check your model’s documentation for all accepted arguments.
bool) —
Whether or not to return the loss only. List[str], optional) —
A list of keys in the output of your model (if it is a dictionary) that should be ignored when
gathering predictions. Returns
Tuple[Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor]]
A tuple with the loss, logits and labels (each being optional).
Perform an evaluation step on model using inputs.
Subclass and override to inject custom behavior.
Sets values in the deepspeed plugin based on the Trainer args
( commit_message: Optional = 'End of training' blocking: bool = True **kwargs )
Parameters
str, optional, defaults to "End of training") —
Message to commit while pushing. bool, optional, defaults to True) —
Whether the function should return only when the git push has finished. Dict[str, Any], optional) —
Additional keyword arguments passed along to create_model_card(). Upload self.model and self.tokenizer to the 🤗 model hub on the repo self.args.hub_model_id.
( callback )
Parameters
type or TrainerCallback) —
A TrainerCallback class or an instance of a TrainerCallback. In the
first case, will remove the first member of that class found in the list of callbacks. Remove a callback from the current list of TrainerCallback.
( split metrics combined = True )
Save metrics into a json file for that split, e.g. train_results.json.
Under distributed environment this is done only for a process with rank 0.
To understand the metrics please read the docstring of log_metrics(). The only difference is that raw unformatted numbers are saved in the current method.
Will save the model, so you can reload it using from_pretrained().
Will only save from the main process.
Saves the Trainer state, since Trainer.save_model saves only the tokenizer with the model
Under distributed environment this is done only for a process with rank 0.
( resume_from_checkpoint: Union = None trial: Union = None ignore_keys_for_eval: Optional = None **kwargs )
Parameters
str or bool, optional) —
If a str, local path to a saved checkpoint as saved by a previous instance of Trainer. If a
bool and equals True, load the last checkpoint in args.output_dir as saved by a previous instance
of Trainer. If present, training will resume from the model/optimizer/scheduler states loaded here. optuna.Trial or Dict[str, Any], optional) —
The trial run or the hyperparameter dictionary for hyperparameter search. List[str], optional) —
A list of keys in the output of your model (if it is a dictionary) that should be ignored when
gathering predictions for evaluation during the training. Dict[str, Any], optional) —
Additional keyword arguments used to hide deprecated arguments Main training entry point.
( model: Module inputs: Dict ) → torch.Tensor
Parameters
nn.Module) —
The model to train. Dict[str, Union[torch.Tensor, Any]]) —
The inputs and targets of the model.
The dictionary will be unpacked before being fed to the model. Most models expect the targets under the
argument labels. Check your model’s documentation for all accepted arguments.
Returns
torch.Tensor
The tensor with training loss on this batch.
Perform a training step on a batch of inputs.
Subclass and override to inject custom behavior.
( model: Union = None args: TrainingArguments = None data_collator: Optional = None train_dataset: Optional = None eval_dataset: Union = None tokenizer: Optional = None model_init: Optional = None compute_metrics: Optional = None callbacks: Optional = None optimizers: Tuple = (None, None) preprocess_logits_for_metrics: Optional = None )
( eval_dataset: Optional = None ignore_keys: Optional = None metric_key_prefix: str = 'eval' **gen_kwargs )
Parameters
Dataset, optional) —
Pass a dataset if you wish to override self.eval_dataset. If it is an Dataset, columns
not accepted by the model.forward() method are automatically removed. It must implement the __len__
method. List[str], optional) —
A list of keys in the output of your model (if it is a dictionary) that should be ignored when
gathering predictions. str, optional, defaults to "eval") —
An optional prefix to be used as the metrics key prefix. For example the metrics “bleu” will be named
“eval_bleu” if the prefix is "eval" (default) int, optional) —
The maximum target length to use when predicting with the generate method. int, optional) —
Number of beams for beam search that will be used when predicting with the generate method. 1 means no
beam search.
gen_kwargs —
Additional generate specific kwargs. Run evaluation and returns metrics.
The calling script will be responsible for providing a method to compute metrics, as they are task-dependent
(pass it to the init compute_metrics argument).
You can also subclass and override this method to inject custom behavior.
( test_dataset: Dataset ignore_keys: Optional = None metric_key_prefix: str = 'test' **gen_kwargs )
Parameters
Dataset) —
Dataset to run the predictions on. If it is a Dataset, columns not accepted by the
model.forward() method are automatically removed. Has to implement the method __len__ List[str], optional) —
A list of keys in the output of your model (if it is a dictionary) that should be ignored when
gathering predictions. str, optional, defaults to "eval") —
An optional prefix to be used as the metrics key prefix. For example the metrics “bleu” will be named
“eval_bleu” if the prefix is "eval" (default) int, optional) —
The maximum target length to use when predicting with the generate method. int, optional) —
Number of beams for beam search that will be used when predicting with the generate method. 1 means no
beam search.
gen_kwargs —
Additional generate specific kwargs. Run prediction and returns predictions and potential metrics.
Depending on the dataset and your use case, your test dataset may contain labels. In that case, this method
will also return metrics, like in evaluate().
If your predictions or labels have different sequence lengths (for instance because you’re doing dynamic padding in a token classification task) the predictions will be padded (on the right) to allow for concatenation into one array. The padding index is -100.
Returns: NamedTuple A namedtuple with the following keys:
np.ndarray): The predictions on test_dataset.np.ndarray, optional): The labels (if the dataset contained some).Dict[str, float], optional): The potential dictionary of metrics (if the dataset contained
labels).( output_dir: str overwrite_output_dir: bool = False do_train: bool = False do_eval: bool = False do_predict: bool = False evaluation_strategy: Union = 'no' prediction_loss_only: bool = False per_device_train_batch_size: int = 8 per_device_eval_batch_size: int = 8 per_gpu_train_batch_size: Optional = None per_gpu_eval_batch_size: Optional = None gradient_accumulation_steps: int = 1 eval_accumulation_steps: Optional = None eval_delay: Optional = 0 learning_rate: float = 5e-05 weight_decay: float = 0.0 adam_beta1: float = 0.9 adam_beta2: float = 0.999 adam_epsilon: float = 1e-08 max_grad_norm: float = 1.0 num_train_epochs: float = 3.0 max_steps: int = -1 lr_scheduler_type: Union = 'linear' lr_scheduler_kwargs: Optional = <factory> warmup_ratio: float = 0.0 warmup_steps: int = 0 log_level: Optional = 'passive' log_level_replica: Optional = 'warning' log_on_each_node: bool = True logging_dir: Optional = None logging_strategy: Union = 'steps' logging_first_step: bool = False logging_steps: float = 500 logging_nan_inf_filter: bool = True save_strategy: Union = 'steps' save_steps: float = 500 save_total_limit: Optional = None save_safetensors: Optional = True save_on_each_node: bool = False save_only_model: bool = False no_cuda: bool = False use_cpu: bool = False use_mps_device: bool = False seed: int = 42 data_seed: Optional = None jit_mode_eval: bool = False use_ipex: bool = False bf16: bool = False fp16: bool = False fp16_opt_level: str = 'O1' half_precision_backend: str = 'auto' bf16_full_eval: bool = False fp16_full_eval: bool = False tf32: Optional = None local_rank: int = -1 ddp_backend: Optional = None tpu_num_cores: Optional = None tpu_metrics_debug: bool = False debug: Union = '' dataloader_drop_last: bool = False eval_steps: Optional = None dataloader_num_workers: int = 0 dataloader_prefetch_factor: int = None past_index: int = -1 run_name: Optional = None disable_tqdm: Optional = None remove_unused_columns: Optional = True label_names: Optional = None load_best_model_at_end: Optional = False metric_for_best_model: Optional = None greater_is_better: Optional = None ignore_data_skip: bool = False fsdp: Union = '' fsdp_min_num_params: int = 0 fsdp_config: Union = None fsdp_transformer_layer_cls_to_wrap: Optional = None accelerator_config: Optional = None deepspeed: Optional = None label_smoothing_factor: float = 0.0 optim: Union = 'adamw_torch' optim_args: Optional = None adafactor: bool = False group_by_length: bool = False length_column_name: Optional = 'length' report_to: Optional = None ddp_find_unused_parameters: Optional = None ddp_bucket_cap_mb: Optional = None ddp_broadcast_buffers: Optional = None dataloader_pin_memory: bool = True dataloader_persistent_workers: bool = False skip_memory_metrics: bool = True use_legacy_prediction_loop: bool = False push_to_hub: bool = False resume_from_checkpoint: Optional = None hub_model_id: Optional = None hub_strategy: Union = 'every_save' hub_token: Optional = None hub_private_repo: bool = False hub_always_push: bool = False gradient_checkpointing: bool = False gradient_checkpointing_kwargs: Optional = None include_inputs_for_metrics: bool = False fp16_backend: str = 'auto' push_to_hub_model_id: Optional = None push_to_hub_organization: Optional = None push_to_hub_token: Optional = None mp_parameters: str = '' auto_find_batch_size: bool = False full_determinism: bool = False torchdynamo: Optional = None ray_scope: Optional = 'last' ddp_timeout: Optional = 1800 torch_compile: bool = False torch_compile_backend: Optional = None torch_compile_mode: Optional = None dispatch_batches: Optional = None split_batches: Optional = None include_tokens_per_second: Optional = False include_num_input_tokens_seen: Optional = False neftune_noise_alpha: Optional = None )
Parameters
str) —
The output directory where the model predictions and checkpoints will be written. bool, optional, defaults to False) —
If True, overwrite the content of the output directory. Use this to continue training if output_dir
points to a checkpoint directory. bool, optional, defaults to False) —
Whether to run training or not. This argument is not directly used by Trainer, it’s intended to be used
by your training/evaluation scripts instead. See the example
scripts for more details. bool, optional) —
Whether to run evaluation on the validation set or not. Will be set to True if evaluation_strategy is
different from "no". This argument is not directly used by Trainer, it’s intended to be used by your
training/evaluation scripts instead. See the example
scripts for more details. bool, optional, defaults to False) —
Whether to run predictions on the test set or not. This argument is not directly used by Trainer, it’s
intended to be used by your training/evaluation scripts instead. See the example
scripts for more details. str or IntervalStrategy, optional, defaults to "no") —
The evaluation strategy to adopt during training. Possible values are:
"no": No evaluation is done during training."steps": Evaluation is done (and logged) every eval_steps."epoch": Evaluation is done at the end of each epoch.bool, optional, defaults to False) —
When performing evaluation and generating predictions, only returns the loss. int, optional, defaults to 8) —
The batch size per GPU/XPU/TPU/MPS/NPU core/CPU for training. int, optional, defaults to 8) —
The batch size per GPU/XPU/TPU/MPS/NPU core/CPU for evaluation. int, optional, defaults to 1) —
Number of updates steps to accumulate the gradients for, before performing a backward/update pass.
When using gradient accumulation, one step is counted as one step with backward pass. Therefore, logging,
evaluation, save will be conducted every gradient_accumulation_steps * xxx_step training examples.
int, optional) —
Number of predictions steps to accumulate the output tensors for, before moving the results to the CPU. If
left unset, the whole predictions are accumulated on GPU/NPU/TPU before being moved to the CPU (faster but
requires more memory). float, optional) —
Number of epochs or steps to wait for before the first evaluation can be performed, depending on the
evaluation_strategy. float, optional, defaults to 5e-5) —
The initial learning rate for AdamW optimizer. float, optional, defaults to 0) —
The weight decay to apply (if not zero) to all layers except all bias and LayerNorm weights in AdamW
optimizer. float, optional, defaults to 0.9) —
The beta1 hyperparameter for the AdamW optimizer. float, optional, defaults to 0.999) —
The beta2 hyperparameter for the AdamW optimizer. float, optional, defaults to 1e-8) —
The epsilon hyperparameter for the AdamW optimizer. float, optional, defaults to 1.0) —
Maximum gradient norm (for gradient clipping). float, optional, defaults to 3.0) —
Total number of training epochs to perform (if not an integer, will perform the decimal part percents of
the last epoch before stopping training). int, optional, defaults to -1) —
If set to a positive number, the total number of training steps to perform. Overrides num_train_epochs.
For a finite dataset, training is reiterated through the dataset (if all data is exhausted) until
max_steps is reached. str or SchedulerType, optional, defaults to "linear") —
The scheduler type to use. See the documentation of SchedulerType for all possible values. float, optional, defaults to 0.0) —
Ratio of total training steps used for a linear warmup from 0 to learning_rate. int, optional, defaults to 0) —
Number of steps used for a linear warmup from 0 to learning_rate. Overrides any effect of warmup_ratio. str, optional, defaults to passive) —
Logger log level to use on the main process. Possible choices are the log levels as strings: ‘debug’,
‘info’, ‘warning’, ‘error’ and ‘critical’, plus a ‘passive’ level which doesn’t set anything and keeps the
current log level for the Transformers library (which will be "warning" by default). str, optional, defaults to "warning") —
Logger log level to use on replicas. Same choices as log_level” bool, optional, defaults to True) —
In multinode distributed training, whether to log using log_level once per node, or only on the main
node. str, optional) —
TensorBoard log directory. Will default to
*output_dir/runs/CURRENT_DATETIME_HOSTNAME*. str or IntervalStrategy, optional, defaults to "steps") —
The logging strategy to adopt during training. Possible values are:
"no": No logging is done during training."epoch": Logging is done at the end of each epoch."steps": Logging is done every logging_steps.bool, optional, defaults to False) —
Whether to log the first global_step or not. int or float, optional, defaults to 500) —
Number of update steps between two logs if logging_strategy="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. bool, optional, defaults to True) —
Whether to filter nan and inf losses for logging. If set to True the loss of every step that is nan
or inf is filtered and the average loss of the current logging window is taken instead.
logging_nan_inf_filter only influences the logging of loss values, it does not change the behavior the
gradient is computed or applied to the model.
str or IntervalStrategy, optional, defaults to "steps") —
The checkpoint save strategy to adopt during training. Possible values are:
"no": No save is done during training."epoch": Save is done at the end of each epoch."steps": Save is done every save_steps.int or float, optional, defaults to 500) —
Number of updates steps before two checkpoint saves if save_strategy="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. int, optional) —
If a value is passed, will limit the total amount of checkpoints. Deletes the older checkpoints in
output_dir. When load_best_model_at_end is enabled, the “best” checkpoint according to
metric_for_best_model will always be retained in addition to the most recent ones. For example, for
save_total_limit=5 and load_best_model_at_end, the four last checkpoints will always be retained
alongside the best model. When save_total_limit=1 and load_best_model_at_end, it is possible that two
checkpoints are saved: the last one and the best one (if they are different). bool, optional, defaults to True) —
Use safetensors saving and loading for state dicts instead of
default torch.load and torch.save. bool, optional, defaults to False) —
When doing multi-node distributed training, whether to save models and checkpoints on each node, or only on
the main one.
This should not be activated when the different nodes use the same storage as the files will be saved with the same names for each node.
bool, optional, defaults to False) —
When checkpointing, whether to only save the model, or also the optimizer, scheduler & rng state.
Note that when this is true, you won’t be able to resume training from checkpoint.
This enables you to save storage by not storing the optimizer, scheduler & rng state.
You can only load the model using from_pretrained with this option set to True. bool, optional, defaults to False) —
Whether or not to use cpu. If set to False, we will use cuda or mps device if available. int, optional, defaults to 42) —
Random seed that will be set at the beginning of training. To ensure reproducibility across runs, use the
~Trainer.model_init function to instantiate the model if it has some randomly initialized parameters. int, optional) —
Random seed to be used with data samplers. If not set, random generators for data sampling will use the
same seed as seed. This can be used to ensure reproducibility of data sampling, independent of the model
seed. bool, optional, defaults to False) —
Whether or not to use PyTorch jit trace for inference. bool, optional, defaults to False) —
Use Intel extension for PyTorch when it is available. IPEX
installation. bool, optional, defaults to False) —
Whether to use bf16 16-bit (mixed) precision training instead of 32-bit training. Requires Ampere or higher
NVIDIA architecture or using CPU (use_cpu) or Ascend NPU. This is an experimental API and it may change. bool, optional, defaults to False) —
Whether to use fp16 16-bit (mixed) precision training instead of 32-bit training. str, optional, defaults to ‘O1’) —
For fp16 training, Apex AMP optimization level selected in [‘O0’, ‘O1’, ‘O2’, and ‘O3’]. See details on
the Apex documentation. str, optional, defaults to "auto") —
This argument is deprecated. Use half_precision_backend instead. str, optional, defaults to "auto") —
The backend to use for mixed precision training. Must be one of "auto", "apex", "cpu_amp". "auto" will
use CPU/CUDA AMP or APEX depending on the PyTorch version detected, while the other choices will force the
requested backend. bool, optional, defaults to False) —
Whether to use full bfloat16 evaluation instead of 32-bit. This will be faster and save memory but can harm
metric values. This is an experimental API and it may change. bool, optional, defaults to False) —
Whether to use full float16 evaluation instead of 32-bit. This will be faster and save memory but can harm
metric values. bool, optional) —
Whether to enable the TF32 mode, available in Ampere and newer GPU architectures. The default value depends
on PyTorch’s version default of torch.backends.cuda.matmul.allow_tf32. For more details please refer to
the TF32 documentation. This is an
experimental API and it may change. int, optional, defaults to -1) —
Rank of the process during distributed training. str, optional) —
The backend to use for distributed training. Must be one of "nccl", "mpi", "ccl", "gloo", "hccl". int, optional) —
When training on TPU, the number of TPU cores (automatically passed by launcher script). bool, optional, defaults to False) —
Whether to drop the last incomplete batch (if the length of the dataset is not divisible by the batch size)
or not. int or float, optional) —
Number of update steps between two evaluations if evaluation_strategy="steps". Will default to the same
value as logging_steps if not set. Should be an integer or a float in range [0,1). If smaller than 1,
will be interpreted as ratio of total training steps. int, optional, defaults to 0) —
Number of subprocesses to use for data loading (PyTorch only). 0 means that the data will be loaded in the
main process. int, optional, defaults to -1) —
Some models like TransformerXL or XLNet can make use of
the past hidden states for their predictions. If this argument is set to a positive int, the Trainer will
use the corresponding output (usually index 2) as the past state and feed it to the model at the next
training step under the keyword argument mems. str, optional) —
A descriptor for the run. Typically used for wandb and
mlflow logging. bool, optional) —
Whether or not to disable the tqdm progress bars and table of metrics produced by
~notebook.NotebookTrainingTracker in Jupyter Notebooks. Will default to True if the logging level is
set to warn or lower (default), False otherwise. bool, optional, defaults to True) —
Whether or not to automatically remove the columns unused by the model forward method. List[str], optional) —
The list of keys in your dictionary of inputs that correspond to the labels.
Will eventually default to the list of argument names accepted by the model that contain the word “label”,
except if the model used is one of the XxxForQuestionAnswering in which case it will also include the
["start_positions", "end_positions"] keys.
bool, optional, defaults to False) —
Whether or not to load the best model found during training at the end of training. When this option is
enabled, the best checkpoint will always be saved. See
save_total_limit
for more.
When set to True, the parameters save_strategy needs to be the same as evaluation_strategy, and in
the case it is “steps”, save_steps must be a round multiple of eval_steps.
str, optional) —
Use in conjunction with load_best_model_at_end to specify the metric to use to compare two different
models. Must be the name of a metric returned by the evaluation with or without the prefix "eval_". Will
default to "loss" if unspecified and load_best_model_at_end=True (to use the evaluation loss).
If you set this value, greater_is_better will default to True. Don’t forget to set it to False if
your metric is better when lower.
bool, optional) —
Use in conjunction with load_best_model_at_end and metric_for_best_model to specify if better models
should have a greater metric or not. Will default to:
True if metric_for_best_model is set to a value that isn’t "loss" or "eval_loss".False if metric_for_best_model is not set, or set to "loss" or "eval_loss".bool, optional, defaults to False) —
When resuming training, whether or not to skip the epochs and batches to get the data loading at the same
stage as in the previous training. If set to True, the training will begin faster (as that skipping step
can take a long time) but will not yield the same results as the interrupted training would have. bool, str or list of FSDPOption, optional, defaults to '') —
Use PyTorch Distributed Parallel Training (in distributed training only).
A list of options along the following:
"full_shard": Shard parameters, gradients and optimizer states."shard_grad_op": Shard optimizer states and gradients."hybrid_shard": Apply FULL_SHARD within a node, and replicate parameters across nodes."hybrid_shard_zero2": Apply SHARD_GRAD_OP within a node, and replicate parameters across nodes."offload": Offload parameters and gradients to CPUs (only compatible with "full_shard" and
"shard_grad_op")."auto_wrap": Automatically recursively wrap layers with FSDP using default_auto_wrap_policy.str or dict, optional) —
Config to be used with fsdp (Pytorch Distributed Parallel Training). The value is either a location of
fsdp json config file (e.g., fsdp_config.json) or an already loaded json file as dict.
A List of config and its options:
min_num_params (int, optional, defaults to 0):
FSDP’s minimum number of parameters for Default Auto Wrapping. (useful only when fsdp field is
passed).
transformer_layer_cls_to_wrap (List[str], optional):
List of transformer layer class names (case-sensitive) to wrap, e.g, BertLayer, GPTJBlock,
T5Block … (useful only when fsdp flag is passed).
backward_prefetch (str, optional)
FSDP’s backward prefetch mode. Controls when to prefetch next set of parameters (useful only when
fsdp field is passed).
A list of options along the following:
"backward_pre" : Prefetches the next set of parameters before the current set of parameter’s
gradient
computation."backward_post" : This prefetches the next set of parameters after the current set of
parameter’s
gradient computation.forward_prefetch (bool, optional, defaults to False)
FSDP’s forward prefetch mode (useful only when fsdp field is passed).
If "True", then FSDP explicitly prefetches the next upcoming all-gather while executing in the
forward pass.
limit_all_gathers (bool, optional, defaults to False)
FSDP’s limit_all_gathers (useful only when fsdp field is passed).
If "True", FSDP explicitly synchronizes the CPU thread to prevent too many in-flight
all-gathers.
use_orig_params (bool, optional, defaults to True)
If "True", allows non-uniform requires_grad during init, which means support for interspersed
frozen and trainable paramteres. Useful in cases such as parameter-efficient fine-tuning. Please
refer this
[blog](https://dev-discuss.pytorch.org/t/rethinking-pytorch-fully-sharded-data-parallel-fsdp-from-first-principles/1019
sync_module_states (bool, optional, defaults to True)
If "True", each individually wrapped FSDP unit will broadcast module parameters from rank 0 to
ensure they are the same across all ranks after initialization
activation_checkpointing (bool, optional, defaults to False):
If "True", activation checkpointing is a technique to reduce memory usage by clearing activations of
certain layers and recomputing them during a backward pass. Effectively, this trades extra
computation time for reduced memory usage.
xla (bool, optional, defaults to False):
Whether to use PyTorch/XLA Fully Sharded Data Parallel Training. This is an experimental feature
and its API may evolve in the future.
xla_fsdp_settings (dict, optional)
The value is a dictionary which stores the XLA FSDP wrapping parameters.
For a complete list of options, please see here.
xla_fsdp_grad_ckpt (bool, optional, defaults to False):
Will use gradient checkpointing over each nested XLA FSDP wrapped layer. This setting can only be
used when the xla flag is set to true, and an auto wrapping policy is specified through
fsdp_min_num_params or fsdp_transformer_layer_cls_to_wrap.
str or dict, optional) —
Use Deepspeed. This is an experimental feature and its API may
evolve in the future. The value is either the location of DeepSpeed json config file (e.g.,
ds_config.json) or an already loaded json file as a dict” str, dict, or AcceleratorConfig, optional) —
Config to be used with the internal Accelerator implementation. The value is either a location of
accelerator json config file (e.g., accelerator_config.json), an already loaded json file as dict,
or an instance of AcceleratorConfig.
A list of config and its options:
bool, optional, defaults to False):
Whether or not the accelerator should split the batches yielded by the dataloaders across the devices. If
True the actual batch size used will be the same on any kind of distributed processes, but it must be a
round multiple of the num_processes you are using. If False, actual batch size used will be the one set
in your script multiplied by the number of processes.bool, optional):
If set to True, the dataloader prepared by the Accelerator is only iterated through on the main process
and then the batches are split and broadcast to each process. Will default to True for DataLoader whose
underlying dataset is an IterableDataset, False otherwise.bool, optional, defaults to True):
If set to True, in cases where the total batch size across all processes does not exactly divide the
dataset, samples at the start of the dataset will be duplicated so the batch can be divided equally among
all workers.bool, optional, defaults to True):
Whether or not use a fully seedable random sampler (accelerate.data_loader.SeedableRandomSampler). Ensures
training results are fully reproducable using a different sampling technique. While seed-to-seed results
may differ, on average the differences are neglible when using multiple different seeds to compare. Should
also be ran with ~utils.set_seed for the best results.float, optional, defaults to 0.0) —
The label smoothing factor to use. Zero means no label smoothing, otherwise the underlying onehot-encoded
labels are changed from 0s and 1s to label_smoothing_factor/num_labels and 1 - label_smoothing_factor + label_smoothing_factor/num_labels respectively. str or list of DebugOption, optional, defaults to "") —
Enable one or more debug features. This is an experimental feature.
Possible options are:
"underflow_overflow": detects overflow in model’s input/outputs and reports the last frames that led to
the event"tpu_metrics_debug": print debug metrics on TPUThe options should be separated by whitespaces.
str or training_args.OptimizerNames, optional, defaults to "adamw_torch") —
The optimizer to use: adamw_hf, adamw_torch, adamw_torch_fused, adamw_apex_fused, adamw_anyprecision or
adafactor. str, optional) —
Optional arguments that are supplied to AnyPrecisionAdamW. bool, optional, defaults to False) —
Whether or not to group together samples of roughly the same length in the training dataset (to minimize
padding applied and be more efficient). Only useful if applying dynamic padding. str, optional, defaults to "length") —
Column name for precomputed lengths. If the column exists, grouping by length will use these values rather
than computing them on train startup. Ignored unless group_by_length is True and the dataset is an
instance of Dataset. str or List[str], optional, defaults to "all") —
The list of integrations to report the results and logs to. Supported platforms are "azure_ml",
"clearml", "codecarbon", "comet_ml", "dagshub", "dvclive", "flyte", "mlflow", "neptune",
"tensorboard", and "wandb". Use "all" to report to all integrations installed, "none" for no
integrations. bool, optional) —
When using distributed training, the value of the flag find_unused_parameters passed to
DistributedDataParallel. Will default to False if gradient checkpointing is used, True otherwise. int, optional) —
When using distributed training, the value of the flag bucket_cap_mb passed to DistributedDataParallel. bool, optional) —
When using distributed training, the value of the flag broadcast_buffers passed to
DistributedDataParallel. Will default to False if gradient checkpointing is used, True otherwise. bool, optional, defaults to True) —
Whether you want to pin memory in data loaders or not. Will default to True. bool, optional, defaults to False) —
If True, the data loader will not shut down the worker processes after a dataset has been consumed once.
This allows to maintain the workers Dataset instances alive. Can potentially speed up training, but will
increase RAM usage. Will default to False. int, optional) —
Number of batches loaded in advance by each worker.
2 means there will be a total of 2 * num_workers batches prefetched across all workers. bool, optional, defaults to True) —
Whether to skip adding of memory profiler reports to metrics. This is skipped by default because it slows
down the training and evaluation speed. bool, optional, defaults to False) —
Whether or not to push the model to the Hub every time the model is saved. If this is activated,
output_dir will begin a git directory synced with the repo (determined by hub_model_id) and the content
will be pushed each time a save is triggered (depending on your save_strategy). Calling
save_model() will also trigger a push.
If output_dir exists, it needs to be a local clone of the repository to which the Trainer will be
pushed.
str, optional) —
The path to a folder with a valid checkpoint for your model. This argument is not directly used by
Trainer, it’s intended to be used by your training/evaluation scripts instead. See the example
scripts for more details. str, optional) —
The name of the repository to keep in sync with the local output_dir. It can be a simple model ID in
which case the model will be pushed in your namespace. Otherwise it should be the whole repository name,
for instance "user_name/model", which allows you to push to an organization you are a member of with
"organization_name/model". Will default to user_name/output_dir_name with output_dir_name being the
name of output_dir.
Will default to the name of output_dir.
str or HubStrategy, optional, defaults to "every_save") —
Defines the scope of what is pushed to the Hub and when. Possible values are:
"end": push the model, its configuration, the tokenizer (if passed along to the Trainer) and a
draft of a model card when the save_model() method is called."every_save": push the model, its configuration, the tokenizer (if passed along to the Trainer) and
a draft of a model card each time there is a model save. The pushes are asynchronous to not block
training, and in case the save are very frequent, a new push is only attempted if the previous one is
finished. A last push is made with the final model at the end of training."checkpoint": like "every_save" but the latest checkpoint is also pushed in a subfolder named
last-checkpoint, allowing you to resume training easily with
trainer.train(resume_from_checkpoint="last-checkpoint")."all_checkpoints": like "checkpoint" but all checkpoints are pushed like they appear in the output
folder (so you will get one checkpoint folder per folder in your final repository)str, optional) —
The token to use to push the model to the Hub. Will default to the token in the cache folder obtained with
huggingface-cli login. bool, optional, defaults to False) —
If True, the Hub repo will be set to private. bool, optional, defaults to False) —
Unless this is True, the Trainer will skip pushing a checkpoint when the previous push is not finished. bool, optional, defaults to False) —
If True, use gradient checkpointing to save memory at the expense of slower backward pass. dict, optional, defaults to None) —
Key word arguments to be passed to the gradient_checkpointing_enable method. bool, optional, defaults to False) —
Whether or not the inputs will be passed to the compute_metrics function. This is intended for metrics
that need inputs, predictions and references for scoring calculation in Metric class. bool, optional, defaults to False) —
Whether to find a batch size that will fit into memory automatically through exponential decay, avoiding
CUDA Out-of-Memory errors. Requires accelerate to be installed (pip install accelerate) bool, optional, defaults to False) —
If True, enable_full_determinism() is called instead of set_seed() to ensure reproducible results in
distributed training. Important: this will negatively impact the performance, so only use it for debugging. str, optional) —
If set, the backend compiler for TorchDynamo. Possible choices are "eager", "aot_eager", "inductor",
"nvfuser", "aot_nvfuser", "aot_cudagraphs", "ofi", "fx2trt", "onnxrt" and "ipex". str, optional, defaults to "last") —
The scope to use when doing hyperparameter search with Ray. By default, "last" will be used. Ray will
then use the last checkpoint of all trials, compare those, and select the best one. However, other options
are also available. See the Ray documentation for
more options. int, optional, defaults to 1800) —
The timeout for torch.distributed.init_process_group calls, used to avoid GPU socket timeouts when
performing slow operations in distributed runnings. Please refer the [PyTorch documentation]
(https://pytorch.org/docs/stable/distributed.html#torch.distributed.init_process_group) for more
information. bool, optional, defaults to False) —
This argument is deprecated.mps device will be used if it is available similar to cuda device. bool, optional, defaults to False) —
Whether or not to compile the model using PyTorch 2.0
torch.compile.
This will use the best defaults for the torch.compile
API.
You can customize the defaults with the argument torch_compile_backend and torch_compile_mode but we
don’t guarantee any of them will work as the support is progressively rolled in in PyTorch.
This flag and the whole compile API is experimental and subject to change in future releases.
str, optional) —
The backend to use in torch.compile. If set to any value, torch_compile will be set to True.
Refer to the PyTorch doc for possible values and note that they may change across PyTorch versions.
This flag is experimental and subject to change in future releases.
str, optional) —
The mode to use in torch.compile. If set to any value, torch_compile will be set to True.
Refer to the PyTorch doc for possible values and note that they may change across PyTorch versions.
This flag is experimental and subject to change in future releases.
bool, optional) —
Whether or not the accelerator should split the batches yielded by the dataloaders across the devices
during distributed training. If
set to True, the actual batch size used will be the same on any kind of distributed processes, but it
must be a
round multiple of the number of processes you are using (such as GPUs).
bool, optional) —
Whether or not to compute the number of tokens per second per device for training speed metrics.
This will iterate over the entire training dataloader once beforehand,
and will slow down the entire process.
bool, optional) —
Whether or not to track the number of input tokens seen throughout training.
May be slower in distributed training as gather operations must be called.
Optional[float]) —
If not None, this will activate NEFTune noise embeddings. This can drastically improve model performance
for instruction fine-tuning. Check out the original paper and the
original code. Support transformers PreTrainedModel and also
PeftModel from peft. TrainingArguments is the subset of the arguments we use in our example scripts which relate to the training loop itself.
Using HfArgumentParser we can turn this class into argparse arguments that can be specified on the command line.
Returns the log level to be used depending on whether this process is the main process of node 0, main process of node non-0, or a non-main process.
For the main process the log level defaults to the logging level set (logging.WARNING if you didn’t do
anything) unless overridden by log_level argument.
For the replica processes the log level defaults to logging.WARNING unless overridden by log_level_replica
argument.
The choice between the main and replica process settings is made according to the return value of should_log.
Get number of steps used for a linear warmup.
( local = True desc = 'work' )
Parameters
bool, optional, defaults to True) —
if True first means process of rank 0 of each node if False first means process of rank 0 of node
rank 0 In multi-node environment with a shared filesystem you most likely will want to use
local=False so that only the main process of the first node will do the processing. If however, the
filesystem is not shared, then the main process of each node will need to do the processing, which is
the default behavior. str, optional, defaults to "work") —
a work description to be used in debug logs A context manager for torch distributed environment where on needs to do something on the main process, while blocking replicas, and when it’s finished releasing the replicas.
One such use is for datasets’s map feature which to be efficient should be run once on the main process,
which upon completion saves a cached version of results and which then automatically gets loaded by the
replicas.
( train_batch_size: int = 8 eval_batch_size: int = 8 drop_last: bool = False num_workers: int = 0 pin_memory: bool = True persistent_workers: bool = False prefetch_factor: Optional = None auto_find_batch_size: bool = False ignore_data_skip: bool = False sampler_seed: Optional = None )
Parameters
bool, optional, defaults to False) —
Whether to drop the last incomplete batch (if the length of the dataset is not divisible by the batch
size) or not. int, optional, defaults to 0) —
Number of subprocesses to use for data loading (PyTorch only). 0 means that the data will be loaded in
the main process. bool, optional, defaults to True) —
Whether you want to pin memory in data loaders or not. Will default to True. bool, optional, defaults to False) —
If True, the data loader will not shut down the worker processes after a dataset has been consumed
once. This allows to maintain the workers Dataset instances alive. Can potentially speed up training,
but will increase RAM usage. Will default to False. int, optional) —
Number of batches loaded in advance by each worker.
2 means there will be a total of 2 * num_workers batches prefetched across all workers. bool, optional, defaults to False) —
Whether to find a batch size that will fit into memory automatically through exponential decay,
avoiding CUDA Out-of-Memory errors. Requires accelerate to be installed (pip install accelerate) bool, optional, defaults to False) —
When resuming training, whether or not to skip the epochs and batches to get the data loading at the
same stage as in the previous training. If set to True, the training will begin faster (as that
skipping step can take a long time) but will not yield the same results as the interrupted training
would have. int, optional) —
Random seed to be used with data samplers. If not set, random generators for data sampling will use the
same seed as self.seed. This can be used to ensure reproducibility of data sampling, independent of
the model seed. A method that regroups all arguments linked to the dataloaders creation.
( strategy: Union = 'no' steps: int = 500 batch_size: int = 8 accumulation_steps: Optional = None delay: Optional = None loss_only: bool = False jit_mode: bool = False )
Parameters
str or IntervalStrategy, optional, defaults to "no") —
The evaluation strategy to adopt during training. Possible values are:
"no": No evaluation is done during training."steps": Evaluation is done (and logged) every steps."epoch": Evaluation is done at the end of each epoch.Setting a strategy different from "no" will set self.do_eval to True.
int, optional, defaults to 500) —
Number of update steps between two evaluations if strategy="steps". int optional, defaults to 8) —
The batch size per device (GPU/TPU core/CPU…) used for evaluation. int, optional) —
Number of predictions steps to accumulate the output tensors for, before moving the results to the CPU.
If left unset, the whole predictions are accumulated on GPU/TPU before being moved to the CPU (faster
but requires more memory). float, optional) —
Number of epochs or steps to wait for before the first evaluation can be performed, depending on the
evaluation_strategy. bool, optional, defaults to False) —
Ignores all outputs except the loss. bool, optional) —
Whether or not to use PyTorch jit trace for inference. A method that regroups all arguments linked to evaluation.
( strategy: Union = 'steps' steps: int = 500 report_to: Union = 'none' level: str = 'passive' first_step: bool = False nan_inf_filter: bool = False on_each_node: bool = False replica_level: str = 'passive' )
Parameters
str or IntervalStrategy, optional, defaults to "steps") —
The logging strategy to adopt during training. Possible values are:
"no": No logging is done during training."epoch": Logging is done at the end of each epoch."steps": Logging is done every logging_steps.int, optional, defaults to 500) —
Number of update steps between two logs if strategy="steps". str, optional, defaults to "passive") —
Logger log level to use on the main process. Possible choices are the log levels as strings: "debug",
"info", "warning", "error" and "critical", plus a "passive" level which doesn’t set anything
and lets the application set the level. str or List[str], optional, defaults to "all") —
The list of integrations to report the results and logs to. Supported platforms are "azure_ml",
"clearml", "codecarbon", "comet_ml", "dagshub", "dvclive", "flyte", "mlflow",
"neptune", "tensorboard", and "wandb". Use "all" to report to all integrations installed,
"none" for no integrations. bool, optional, defaults to False) —
Whether to log and evaluate the first global_step or not. bool, optional, defaults to True) —
Whether to filter nan and inf losses for logging. If set to True the loss of every step that is
nan or inf is filtered and the average loss of the current logging window is taken instead.
nan_inf_filter only influences the logging of loss values, it does not change the behavior the
gradient is computed or applied to the model.
bool, optional, defaults to True) —
In multinode distributed training, whether to log using log_level once per node, or only on the main
node. str, optional, defaults to "passive") —
Logger log level to use on replicas. Same choices as log_level A method that regroups all arguments linked to logging.
( name: Union = 'linear' num_epochs: float = 3.0 max_steps: int = -1 warmup_ratio: float = 0 warmup_steps: int = 0 )
Parameters
str or SchedulerType, optional, defaults to "linear") —
The scheduler type to use. See the documentation of SchedulerType for all possible values. float, optional, defaults to 3.0) —
Total number of training epochs to perform (if not an integer, will perform the decimal part percents
of the last epoch before stopping training). int, optional, defaults to -1) —
If set to a positive number, the total number of training steps to perform. Overrides num_train_epochs.
For a finite dataset, training is reiterated through the dataset (if all data is exhausted) until
max_steps is reached. float, optional, defaults to 0.0) —
Ratio of total training steps used for a linear warmup from 0 to learning_rate. int, optional, defaults to 0) —
Number of steps used for a linear warmup from 0 to learning_rate. Overrides any effect of
warmup_ratio. A method that regroups all arguments linked to the learning rate scheduler and its hyperparameters.
( name: Union = 'adamw_torch' learning_rate: float = 5e-05 weight_decay: float = 0 beta1: float = 0.9 beta2: float = 0.999 epsilon: float = 1e-08 args: Optional = None )
Parameters
str or training_args.OptimizerNames, optional, defaults to "adamw_torch") —
The optimizer to use: "adamw_hf", "adamw_torch", "adamw_torch_fused", "adamw_apex_fused",
"adamw_anyprecision" or "adafactor". float, optional, defaults to 5e-5) —
The initial learning rate. float, optional, defaults to 0) —
The weight decay to apply (if not zero) to all layers except all bias and LayerNorm weights. float, optional, defaults to 0.9) —
The beta1 hyperparameter for the adam optimizer or its variants. float, optional, defaults to 0.999) —
The beta2 hyperparameter for the adam optimizer or its variants. float, optional, defaults to 1e-8) —
The epsilon hyperparameter for the adam optimizer or its variants. str, optional) —
Optional arguments that are supplied to AnyPrecisionAdamW (only useful when
optim="adamw_anyprecision"). A method that regroups all arguments linked to the optimizer and its hyperparameters.
( model_id: str strategy: Union = 'every_save' token: Optional = None private_repo: bool = False always_push: bool = False )
Parameters
str) —
The name of the repository to keep in sync with the local output_dir. It can be a simple model ID in
which case the model will be pushed in your namespace. Otherwise it should be the whole repository
name, for instance "user_name/model", which allows you to push to an organization you are a member of
with "organization_name/model". str or HubStrategy, optional, defaults to "every_save") —
Defines the scope of what is pushed to the Hub and when. Possible values are:
"end": push the model, its configuration, the tokenizer (if passed along to the Trainer) and a
draft of a model card when the save_model() method is called."every_save": push the model, its configuration, the tokenizer (if passed along to the Trainer)
and
a draft of a model card each time there is a model save. The pushes are asynchronous to not block
training, and in case the save are very frequent, a new push is only attempted if the previous one is
finished. A last push is made with the final model at the end of training."checkpoint": like "every_save" but the latest checkpoint is also pushed in a subfolder named
last-checkpoint, allowing you to resume training easily with
trainer.train(resume_from_checkpoint="last-checkpoint")."all_checkpoints": like "checkpoint" but all checkpoints are pushed like they appear in the
output
folder (so you will get one checkpoint folder per folder in your final repository)str, optional) —
The token to use to push the model to the Hub. Will default to the token in the cache folder obtained
with huggingface-cli login. bool, optional, defaults to False) —
If True, the Hub repo will be set to private. bool, optional, defaults to False) —
Unless this is True, the Trainer will skip pushing a checkpoint when the previous push is not
finished. A method that regroups all arguments linked to synchronizing checkpoints with the Hub.
Calling this method will set self.push_to_hub to True, which means the output_dir will begin a git
directory synced with the repo (determined by model_id) and the content will be pushed each time a save is
triggered (depending onself.save_strategy). Calling save_model() will also trigger a push.
( strategy: Union = 'steps' steps: int = 500 total_limit: Optional = None on_each_node: bool = False )
Parameters
str or IntervalStrategy, optional, defaults to "steps") —
The checkpoint save strategy to adopt during training. Possible values are:
"no": No save is done during training."epoch": Save is done at the end of each epoch."steps": Save is done every save_steps.int, optional, defaults to 500) —
Number of updates steps before two checkpoint saves if strategy="steps". int, optional) —
If a value is passed, will limit the total amount of checkpoints. Deletes the older checkpoints in
output_dir. bool, optional, defaults to False) —
When doing multi-node distributed training, whether to save models and checkpoints on each node, or
only on the main one.
This should not be activated when the different nodes use the same storage as the files will be saved with the same names for each node.
A method that regroups all arguments linked to checkpoint saving.
( batch_size: int = 8 loss_only: bool = False jit_mode: bool = False )
A method that regroups all basic arguments linked to testing on a held-out dataset.
Calling this method will automatically set self.do_predict to True.
( learning_rate: float = 5e-05 batch_size: int = 8 weight_decay: float = 0 num_epochs: float = 3 max_steps: int = -1 gradient_accumulation_steps: int = 1 seed: int = 42 gradient_checkpointing: bool = False )
Parameters
float, optional, defaults to 5e-5) —
The initial learning rate for the optimizer. int optional, defaults to 8) —
The batch size per device (GPU/TPU core/CPU…) used for training. float, optional, defaults to 0) —
The weight decay to apply (if not zero) to all layers except all bias and LayerNorm weights in the
optimizer. float, optional, defaults to 3.0) —
Total number of training epochs to perform (if not an integer, will perform the decimal part percents
of the last epoch before stopping training). int, optional, defaults to -1) —
If set to a positive number, the total number of training steps to perform. Overrides num_train_epochs.
For a finite dataset, training is reiterated through the dataset (if all data is exhausted) until
max_steps is reached. int, optional, defaults to 1) —
Number of updates steps to accumulate the gradients for, before performing a backward/update pass.
When using gradient accumulation, one step is counted as one step with backward pass. Therefore,
logging, evaluation, save will be conducted every gradient_accumulation_steps * xxx_step training
examples.
int, optional, defaults to 42) —
Random seed that will be set at the beginning of training. To ensure reproducibility across runs, use
the ~Trainer.model_init function to instantiate the model if it has some randomly initialized
parameters. bool, optional, defaults to False) —
If True, use gradient checkpointing to save memory at the expense of slower backward pass. A method that regroups all basic arguments linked to the training.
Calling this method will automatically set self.do_train to True.
Serializes this instance while replace Enum by their values (for JSON serialization support). It obfuscates
the token values by removing their value.
Serializes this instance to a JSON string.
Sanitized serialization to use with TensorBoard’s hparams
( output_dir: str overwrite_output_dir: bool = False do_train: bool = False do_eval: bool = False do_predict: bool = False evaluation_strategy: Union = 'no' prediction_loss_only: bool = False per_device_train_batch_size: int = 8 per_device_eval_batch_size: int = 8 per_gpu_train_batch_size: Optional = None per_gpu_eval_batch_size: Optional = None gradient_accumulation_steps: int = 1 eval_accumulation_steps: Optional = None eval_delay: Optional = 0 learning_rate: float = 5e-05 weight_decay: float = 0.0 adam_beta1: float = 0.9 adam_beta2: float = 0.999 adam_epsilon: float = 1e-08 max_grad_norm: float = 1.0 num_train_epochs: float = 3.0 max_steps: int = -1 lr_scheduler_type: Union = 'linear' lr_scheduler_kwargs: Optional = <factory> warmup_ratio: float = 0.0 warmup_steps: int = 0 log_level: Optional = 'passive' log_level_replica: Optional = 'warning' log_on_each_node: bool = True logging_dir: Optional = None logging_strategy: Union = 'steps' logging_first_step: bool = False logging_steps: float = 500 logging_nan_inf_filter: bool = True save_strategy: Union = 'steps' save_steps: float = 500 save_total_limit: Optional = None save_safetensors: Optional = True save_on_each_node: bool = False save_only_model: bool = False no_cuda: bool = False use_cpu: bool = False use_mps_device: bool = False seed: int = 42 data_seed: Optional = None jit_mode_eval: bool = False use_ipex: bool = False bf16: bool = False fp16: bool = False fp16_opt_level: str = 'O1' half_precision_backend: str = 'auto' bf16_full_eval: bool = False fp16_full_eval: bool = False tf32: Optional = None local_rank: int = -1 ddp_backend: Optional = None tpu_num_cores: Optional = None tpu_metrics_debug: bool = False debug: Union = '' dataloader_drop_last: bool = False eval_steps: Optional = None dataloader_num_workers: int = 0 dataloader_prefetch_factor: int = None past_index: int = -1 run_name: Optional = None disable_tqdm: Optional = None remove_unused_columns: Optional = True label_names: Optional = None load_best_model_at_end: Optional = False metric_for_best_model: Optional = None greater_is_better: Optional = None ignore_data_skip: bool = False fsdp: Union = '' fsdp_min_num_params: int = 0 fsdp_config: Union = None fsdp_transformer_layer_cls_to_wrap: Optional = None accelerator_config: Optional = None deepspeed: Optional = None label_smoothing_factor: float = 0.0 optim: Union = 'adamw_torch' optim_args: Optional = None adafactor: bool = False group_by_length: bool = False length_column_name: Optional = 'length' report_to: Optional = None ddp_find_unused_parameters: Optional = None ddp_bucket_cap_mb: Optional = None ddp_broadcast_buffers: Optional = None dataloader_pin_memory: bool = True dataloader_persistent_workers: bool = False skip_memory_metrics: bool = True use_legacy_prediction_loop: bool = False push_to_hub: bool = False resume_from_checkpoint: Optional = None hub_model_id: Optional = None hub_strategy: Union = 'every_save' hub_token: Optional = None hub_private_repo: bool = False hub_always_push: bool = False gradient_checkpointing: bool = False gradient_checkpointing_kwargs: Optional = None include_inputs_for_metrics: bool = False fp16_backend: str = 'auto' push_to_hub_model_id: Optional = None push_to_hub_organization: Optional = None push_to_hub_token: Optional = None mp_parameters: str = '' auto_find_batch_size: bool = False full_determinism: bool = False torchdynamo: Optional = None ray_scope: Optional = 'last' ddp_timeout: Optional = 1800 torch_compile: bool = False torch_compile_backend: Optional = None torch_compile_mode: Optional = None dispatch_batches: Optional = None split_batches: Optional = None include_tokens_per_second: Optional = False include_num_input_tokens_seen: Optional = False neftune_noise_alpha: Optional = None sortish_sampler: bool = False predict_with_generate: bool = False generation_max_length: Optional = None generation_num_beams: Optional = None generation_config: Union = None )
Parameters
str) —
The output directory where the model predictions and checkpoints will be written. bool, optional, defaults to False) —
If True, overwrite the content of the output directory. Use this to continue training if output_dir
points to a checkpoint directory. bool, optional, defaults to False) —
Whether to run training or not. This argument is not directly used by Trainer, it’s intended to be used
by your training/evaluation scripts instead. See the example
scripts for more details. bool, optional) —
Whether to run evaluation on the validation set or not. Will be set to True if evaluation_strategy is
different from "no". This argument is not directly used by Trainer, it’s intended to be used by your
training/evaluation scripts instead. See the example
scripts for more details. bool, optional, defaults to False) —
Whether to run predictions on the test set or not. This argument is not directly used by Trainer, it’s
intended to be used by your training/evaluation scripts instead. See the example
scripts for more details. str or IntervalStrategy, optional, defaults to "no") —
The evaluation strategy to adopt during training. Possible values are:
"no": No evaluation is done during training."steps": Evaluation is done (and logged) every eval_steps."epoch": Evaluation is done at the end of each epoch.bool, optional, defaults to False) —
When performing evaluation and generating predictions, only returns the loss. int, optional, defaults to 8) —
The batch size per GPU/XPU/TPU/MPS/NPU core/CPU for training. int, optional, defaults to 8) —
The batch size per GPU/XPU/TPU/MPS/NPU core/CPU for evaluation. int, optional, defaults to 1) —
Number of updates steps to accumulate the gradients for, before performing a backward/update pass.
When using gradient accumulation, one step is counted as one step with backward pass. Therefore, logging,
evaluation, save will be conducted every gradient_accumulation_steps * xxx_step training examples.
int, optional) —
Number of predictions steps to accumulate the output tensors for, before moving the results to the CPU. If
left unset, the whole predictions are accumulated on GPU/NPU/TPU before being moved to the CPU (faster but
requires more memory). float, optional) —
Number of epochs or steps to wait for before the first evaluation can be performed, depending on the
evaluation_strategy. float, optional, defaults to 5e-5) —
The initial learning rate for AdamW optimizer. float, optional, defaults to 0) —
The weight decay to apply (if not zero) to all layers except all bias and LayerNorm weights in AdamW
optimizer. float, optional, defaults to 0.9) —
The beta1 hyperparameter for the AdamW optimizer. float, optional, defaults to 0.999) —
The beta2 hyperparameter for the AdamW optimizer. float, optional, defaults to 1e-8) —
The epsilon hyperparameter for the AdamW optimizer. float, optional, defaults to 1.0) —
Maximum gradient norm (for gradient clipping). float, optional, defaults to 3.0) —
Total number of training epochs to perform (if not an integer, will perform the decimal part percents of
the last epoch before stopping training). int, optional, defaults to -1) —
If set to a positive number, the total number of training steps to perform. Overrides num_train_epochs.
For a finite dataset, training is reiterated through the dataset (if all data is exhausted) until
max_steps is reached. str or SchedulerType, optional, defaults to "linear") —
The scheduler type to use. See the documentation of SchedulerType for all possible values. float, optional, defaults to 0.0) —
Ratio of total training steps used for a linear warmup from 0 to learning_rate. int, optional, defaults to 0) —
Number of steps used for a linear warmup from 0 to learning_rate. Overrides any effect of warmup_ratio. str, optional, defaults to passive) —
Logger log level to use on the main process. Possible choices are the log levels as strings: ‘debug’,
‘info’, ‘warning’, ‘error’ and ‘critical’, plus a ‘passive’ level which doesn’t set anything and keeps the
current log level for the Transformers library (which will be "warning" by default). str, optional, defaults to "warning") —
Logger log level to use on replicas. Same choices as log_level” bool, optional, defaults to True) —
In multinode distributed training, whether to log using log_level once per node, or only on the main
node. str, optional) —
TensorBoard log directory. Will default to
*output_dir/runs/CURRENT_DATETIME_HOSTNAME*. str or IntervalStrategy, optional, defaults to "steps") —
The logging strategy to adopt during training. Possible values are:
"no": No logging is done during training."epoch": Logging is done at the end of each epoch."steps": Logging is done every logging_steps.bool, optional, defaults to False) —
Whether to log the first global_step or not. int or float, optional, defaults to 500) —
Number of update steps between two logs if logging_strategy="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. bool, optional, defaults to True) —
Whether to filter nan and inf losses for logging. If set to True the loss of every step that is nan
or inf is filtered and the average loss of the current logging window is taken instead.
logging_nan_inf_filter only influences the logging of loss values, it does not change the behavior the
gradient is computed or applied to the model.
str or IntervalStrategy, optional, defaults to "steps") —
The checkpoint save strategy to adopt during training. Possible values are:
"no": No save is done during training."epoch": Save is done at the end of each epoch."steps": Save is done every save_steps.int or float, optional, defaults to 500) —
Number of updates steps before two checkpoint saves if save_strategy="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. int, optional) —
If a value is passed, will limit the total amount of checkpoints. Deletes the older checkpoints in
output_dir. When load_best_model_at_end is enabled, the “best” checkpoint according to
metric_for_best_model will always be retained in addition to the most recent ones. For example, for
save_total_limit=5 and load_best_model_at_end, the four last checkpoints will always be retained
alongside the best model. When save_total_limit=1 and load_best_model_at_end, it is possible that two
checkpoints are saved: the last one and the best one (if they are different). bool, optional, defaults to True) —
Use safetensors saving and loading for state dicts instead of
default torch.load and torch.save. bool, optional, defaults to False) —
When doing multi-node distributed training, whether to save models and checkpoints on each node, or only on
the main one.
This should not be activated when the different nodes use the same storage as the files will be saved with the same names for each node.
bool, optional, defaults to False) —
When checkpointing, whether to only save the model, or also the optimizer, scheduler & rng state.
Note that when this is true, you won’t be able to resume training from checkpoint.
This enables you to save storage by not storing the optimizer, scheduler & rng state.
You can only load the model using from_pretrained with this option set to True. bool, optional, defaults to False) —
Whether or not to use cpu. If set to False, we will use cuda or mps device if available. int, optional, defaults to 42) —
Random seed that will be set at the beginning of training. To ensure reproducibility across runs, use the
~Trainer.model_init function to instantiate the model if it has some randomly initialized parameters. int, optional) —
Random seed to be used with data samplers. If not set, random generators for data sampling will use the
same seed as seed. This can be used to ensure reproducibility of data sampling, independent of the model
seed. bool, optional, defaults to False) —
Whether or not to use PyTorch jit trace for inference. bool, optional, defaults to False) —
Use Intel extension for PyTorch when it is available. IPEX
installation. bool, optional, defaults to False) —
Whether to use bf16 16-bit (mixed) precision training instead of 32-bit training. Requires Ampere or higher
NVIDIA architecture or using CPU (use_cpu) or Ascend NPU. This is an experimental API and it may change. bool, optional, defaults to False) —
Whether to use fp16 16-bit (mixed) precision training instead of 32-bit training. str, optional, defaults to ‘O1’) —
For fp16 training, Apex AMP optimization level selected in [‘O0’, ‘O1’, ‘O2’, and ‘O3’]. See details on
the Apex documentation. str, optional, defaults to "auto") —
This argument is deprecated. Use half_precision_backend instead. str, optional, defaults to "auto") —
The backend to use for mixed precision training. Must be one of "auto", "apex", "cpu_amp". "auto" will
use CPU/CUDA AMP or APEX depending on the PyTorch version detected, while the other choices will force the
requested backend. bool, optional, defaults to False) —
Whether to use full bfloat16 evaluation instead of 32-bit. This will be faster and save memory but can harm
metric values. This is an experimental API and it may change. bool, optional, defaults to False) —
Whether to use full float16 evaluation instead of 32-bit. This will be faster and save memory but can harm
metric values. bool, optional) —
Whether to enable the TF32 mode, available in Ampere and newer GPU architectures. The default value depends
on PyTorch’s version default of torch.backends.cuda.matmul.allow_tf32. For more details please refer to
the TF32 documentation. This is an
experimental API and it may change. int, optional, defaults to -1) —
Rank of the process during distributed training. str, optional) —
The backend to use for distributed training. Must be one of "nccl", "mpi", "ccl", "gloo", "hccl". int, optional) —
When training on TPU, the number of TPU cores (automatically passed by launcher script). bool, optional, defaults to False) —
Whether to drop the last incomplete batch (if the length of the dataset is not divisible by the batch size)
or not. int or float, optional) —
Number of update steps between two evaluations if evaluation_strategy="steps". Will default to the same
value as logging_steps if not set. Should be an integer or a float in range [0,1). If smaller than 1,
will be interpreted as ratio of total training steps. int, optional, defaults to 0) —
Number of subprocesses to use for data loading (PyTorch only). 0 means that the data will be loaded in the
main process. int, optional, defaults to -1) —
Some models like TransformerXL or XLNet can make use of
the past hidden states for their predictions. If this argument is set to a positive int, the Trainer will
use the corresponding output (usually index 2) as the past state and feed it to the model at the next
training step under the keyword argument mems. str, optional) —
A descriptor for the run. Typically used for wandb and
mlflow logging. bool, optional) —
Whether or not to disable the tqdm progress bars and table of metrics produced by
~notebook.NotebookTrainingTracker in Jupyter Notebooks. Will default to True if the logging level is
set to warn or lower (default), False otherwise. bool, optional, defaults to True) —
Whether or not to automatically remove the columns unused by the model forward method. List[str], optional) —
The list of keys in your dictionary of inputs that correspond to the labels.
Will eventually default to the list of argument names accepted by the model that contain the word “label”,
except if the model used is one of the XxxForQuestionAnswering in which case it will also include the
["start_positions", "end_positions"] keys.
bool, optional, defaults to False) —
Whether or not to load the best model found during training at the end of training. When this option is
enabled, the best checkpoint will always be saved. See
save_total_limit
for more.
When set to True, the parameters save_strategy needs to be the same as evaluation_strategy, and in
the case it is “steps”, save_steps must be a round multiple of eval_steps.
str, optional) —
Use in conjunction with load_best_model_at_end to specify the metric to use to compare two different
models. Must be the name of a metric returned by the evaluation with or without the prefix "eval_". Will
default to "loss" if unspecified and load_best_model_at_end=True (to use the evaluation loss).
If you set this value, greater_is_better will default to True. Don’t forget to set it to False if
your metric is better when lower.
bool, optional) —
Use in conjunction with load_best_model_at_end and metric_for_best_model to specify if better models
should have a greater metric or not. Will default to:
True if metric_for_best_model is set to a value that isn’t "loss" or "eval_loss".False if metric_for_best_model is not set, or set to "loss" or "eval_loss".bool, optional, defaults to False) —
When resuming training, whether or not to skip the epochs and batches to get the data loading at the same
stage as in the previous training. If set to True, the training will begin faster (as that skipping step
can take a long time) but will not yield the same results as the interrupted training would have. bool, str or list of FSDPOption, optional, defaults to '') —
Use PyTorch Distributed Parallel Training (in distributed training only).
A list of options along the following:
"full_shard": Shard parameters, gradients and optimizer states."shard_grad_op": Shard optimizer states and gradients."hybrid_shard": Apply FULL_SHARD within a node, and replicate parameters across nodes."hybrid_shard_zero2": Apply SHARD_GRAD_OP within a node, and replicate parameters across nodes."offload": Offload parameters and gradients to CPUs (only compatible with "full_shard" and
"shard_grad_op")."auto_wrap": Automatically recursively wrap layers with FSDP using default_auto_wrap_policy.str or dict, optional) —
Config to be used with fsdp (Pytorch Distributed Parallel Training). The value is either a location of
fsdp json config file (e.g., fsdp_config.json) or an already loaded json file as dict.
A List of config and its options:
min_num_params (int, optional, defaults to 0):
FSDP’s minimum number of parameters for Default Auto Wrapping. (useful only when fsdp field is
passed).
transformer_layer_cls_to_wrap (List[str], optional):
List of transformer layer class names (case-sensitive) to wrap, e.g, BertLayer, GPTJBlock,
T5Block … (useful only when fsdp flag is passed).
backward_prefetch (str, optional)
FSDP’s backward prefetch mode. Controls when to prefetch next set of parameters (useful only when
fsdp field is passed).
A list of options along the following:
"backward_pre" : Prefetches the next set of parameters before the current set of parameter’s
gradient
computation."backward_post" : This prefetches the next set of parameters after the current set of
parameter’s
gradient computation.forward_prefetch (bool, optional, defaults to False)
FSDP’s forward prefetch mode (useful only when fsdp field is passed).
If "True", then FSDP explicitly prefetches the next upcoming all-gather while executing in the
forward pass.
limit_all_gathers (bool, optional, defaults to False)
FSDP’s limit_all_gathers (useful only when fsdp field is passed).
If "True", FSDP explicitly synchronizes the CPU thread to prevent too many in-flight
all-gathers.
use_orig_params (bool, optional, defaults to True)
If "True", allows non-uniform requires_grad during init, which means support for interspersed
frozen and trainable paramteres. Useful in cases such as parameter-efficient fine-tuning. Please
refer this
[blog](https://dev-discuss.pytorch.org/t/rethinking-pytorch-fully-sharded-data-parallel-fsdp-from-first-principles/1019
sync_module_states (bool, optional, defaults to True)
If "True", each individually wrapped FSDP unit will broadcast module parameters from rank 0 to
ensure they are the same across all ranks after initialization
activation_checkpointing (bool, optional, defaults to False):
If "True", activation checkpointing is a technique to reduce memory usage by clearing activations of
certain layers and recomputing them during a backward pass. Effectively, this trades extra
computation time for reduced memory usage.
xla (bool, optional, defaults to False):
Whether to use PyTorch/XLA Fully Sharded Data Parallel Training. This is an experimental feature
and its API may evolve in the future.
xla_fsdp_settings (dict, optional)
The value is a dictionary which stores the XLA FSDP wrapping parameters.
For a complete list of options, please see here.
xla_fsdp_grad_ckpt (bool, optional, defaults to False):
Will use gradient checkpointing over each nested XLA FSDP wrapped layer. This setting can only be
used when the xla flag is set to true, and an auto wrapping policy is specified through
fsdp_min_num_params or fsdp_transformer_layer_cls_to_wrap.
str or dict, optional) —
Use Deepspeed. This is an experimental feature and its API may
evolve in the future. The value is either the location of DeepSpeed json config file (e.g.,
ds_config.json) or an already loaded json file as a dict” str, dict, or AcceleratorConfig, optional) —
Config to be used with the internal Accelerator implementation. The value is either a location of
accelerator json config file (e.g., accelerator_config.json), an already loaded json file as dict,
or an instance of AcceleratorConfig.
A list of config and its options:
bool, optional, defaults to False):
Whether or not the accelerator should split the batches yielded by the dataloaders across the devices. If
True the actual batch size used will be the same on any kind of distributed processes, but it must be a
round multiple of the num_processes you are using. If False, actual batch size used will be the one set
in your script multiplied by the number of processes.bool, optional):
If set to True, the dataloader prepared by the Accelerator is only iterated through on the main process
and then the batches are split and broadcast to each process. Will default to True for DataLoader whose
underlying dataset is an IterableDataset, False otherwise.bool, optional, defaults to True):
If set to True, in cases where the total batch size across all processes does not exactly divide the
dataset, samples at the start of the dataset will be duplicated so the batch can be divided equally among
all workers.bool, optional, defaults to True):
Whether or not use a fully seedable random sampler (accelerate.data_loader.SeedableRandomSampler). Ensures
training results are fully reproducable using a different sampling technique. While seed-to-seed results
may differ, on average the differences are neglible when using multiple different seeds to compare. Should
also be ran with ~utils.set_seed for the best results.float, optional, defaults to 0.0) —
The label smoothing factor to use. Zero means no label smoothing, otherwise the underlying onehot-encoded
labels are changed from 0s and 1s to label_smoothing_factor/num_labels and 1 - label_smoothing_factor + label_smoothing_factor/num_labels respectively. str or list of DebugOption, optional, defaults to "") —
Enable one or more debug features. This is an experimental feature.
Possible options are:
"underflow_overflow": detects overflow in model’s input/outputs and reports the last frames that led to
the event"tpu_metrics_debug": print debug metrics on TPUThe options should be separated by whitespaces.
str or training_args.OptimizerNames, optional, defaults to "adamw_torch") —
The optimizer to use: adamw_hf, adamw_torch, adamw_torch_fused, adamw_apex_fused, adamw_anyprecision or
adafactor. str, optional) —
Optional arguments that are supplied to AnyPrecisionAdamW. bool, optional, defaults to False) —
Whether or not to group together samples of roughly the same length in the training dataset (to minimize
padding applied and be more efficient). Only useful if applying dynamic padding. str, optional, defaults to "length") —
Column name for precomputed lengths. If the column exists, grouping by length will use these values rather
than computing them on train startup. Ignored unless group_by_length is True and the dataset is an
instance of Dataset. str or List[str], optional, defaults to "all") —
The list of integrations to report the results and logs to. Supported platforms are "azure_ml",
"clearml", "codecarbon", "comet_ml", "dagshub", "dvclive", "flyte", "mlflow", "neptune",
"tensorboard", and "wandb". Use "all" to report to all integrations installed, "none" for no
integrations. bool, optional) —
When using distributed training, the value of the flag find_unused_parameters passed to
DistributedDataParallel. Will default to False if gradient checkpointing is used, True otherwise. int, optional) —
When using distributed training, the value of the flag bucket_cap_mb passed to DistributedDataParallel. bool, optional) —
When using distributed training, the value of the flag broadcast_buffers passed to
DistributedDataParallel. Will default to False if gradient checkpointing is used, True otherwise. bool, optional, defaults to True) —
Whether you want to pin memory in data loaders or not. Will default to True. bool, optional, defaults to False) —
If True, the data loader will not shut down the worker processes after a dataset has been consumed once.
This allows to maintain the workers Dataset instances alive. Can potentially speed up training, but will
increase RAM usage. Will default to False. int, optional) —
Number of batches loaded in advance by each worker.
2 means there will be a total of 2 * num_workers batches prefetched across all workers. bool, optional, defaults to True) —
Whether to skip adding of memory profiler reports to metrics. This is skipped by default because it slows
down the training and evaluation speed. bool, optional, defaults to False) —
Whether or not to push the model to the Hub every time the model is saved. If this is activated,
output_dir will begin a git directory synced with the repo (determined by hub_model_id) and the content
will be pushed each time a save is triggered (depending on your save_strategy). Calling
save_model() will also trigger a push.
If output_dir exists, it needs to be a local clone of the repository to which the Trainer will be
pushed.
str, optional) —
The path to a folder with a valid checkpoint for your model. This argument is not directly used by
Trainer, it’s intended to be used by your training/evaluation scripts instead. See the example
scripts for more details. str, optional) —
The name of the repository to keep in sync with the local output_dir. It can be a simple model ID in
which case the model will be pushed in your namespace. Otherwise it should be the whole repository name,
for instance "user_name/model", which allows you to push to an organization you are a member of with
"organization_name/model". Will default to user_name/output_dir_name with output_dir_name being the
name of output_dir.
Will default to the name of output_dir.
str or HubStrategy, optional, defaults to "every_save") —
Defines the scope of what is pushed to the Hub and when. Possible values are:
"end": push the model, its configuration, the tokenizer (if passed along to the Trainer) and a
draft of a model card when the save_model() method is called."every_save": push the model, its configuration, the tokenizer (if passed along to the Trainer) and
a draft of a model card each time there is a model save. The pushes are asynchronous to not block
training, and in case the save are very frequent, a new push is only attempted if the previous one is
finished. A last push is made with the final model at the end of training."checkpoint": like "every_save" but the latest checkpoint is also pushed in a subfolder named
last-checkpoint, allowing you to resume training easily with
trainer.train(resume_from_checkpoint="last-checkpoint")."all_checkpoints": like "checkpoint" but all checkpoints are pushed like they appear in the output
folder (so you will get one checkpoint folder per folder in your final repository)str, optional) —
The token to use to push the model to the Hub. Will default to the token in the cache folder obtained with
huggingface-cli login. bool, optional, defaults to False) —
If True, the Hub repo will be set to private. bool, optional, defaults to False) —
Unless this is True, the Trainer will skip pushing a checkpoint when the previous push is not finished. bool, optional, defaults to False) —
If True, use gradient checkpointing to save memory at the expense of slower backward pass. dict, optional, defaults to None) —
Key word arguments to be passed to the gradient_checkpointing_enable method. bool, optional, defaults to False) —
Whether or not the inputs will be passed to the compute_metrics function. This is intended for metrics
that need inputs, predictions and references for scoring calculation in Metric class. bool, optional, defaults to False) —
Whether to find a batch size that will fit into memory automatically through exponential decay, avoiding
CUDA Out-of-Memory errors. Requires accelerate to be installed (pip install accelerate) bool, optional, defaults to False) —
If True, enable_full_determinism() is called instead of set_seed() to ensure reproducible results in
distributed training. Important: this will negatively impact the performance, so only use it for debugging. str, optional) —
If set, the backend compiler for TorchDynamo. Possible choices are "eager", "aot_eager", "inductor",
"nvfuser", "aot_nvfuser", "aot_cudagraphs", "ofi", "fx2trt", "onnxrt" and "ipex". str, optional, defaults to "last") —
The scope to use when doing hyperparameter search with Ray. By default, "last" will be used. Ray will
then use the last checkpoint of all trials, compare those, and select the best one. However, other options
are also available. See the Ray documentation for
more options. int, optional, defaults to 1800) —
The timeout for torch.distributed.init_process_group calls, used to avoid GPU socket timeouts when
performing slow operations in distributed runnings. Please refer the [PyTorch documentation]
(https://pytorch.org/docs/stable/distributed.html#torch.distributed.init_process_group) for more
information. bool, optional, defaults to False) —
This argument is deprecated.mps device will be used if it is available similar to cuda device. bool, optional, defaults to False) —
Whether or not to compile the model using PyTorch 2.0
torch.compile.
This will use the best defaults for the torch.compile
API.
You can customize the defaults with the argument torch_compile_backend and torch_compile_mode but we
don’t guarantee any of them will work as the support is progressively rolled in in PyTorch.
This flag and the whole compile API is experimental and subject to change in future releases.
str, optional) —
The backend to use in torch.compile. If set to any value, torch_compile will be set to True.
Refer to the PyTorch doc for possible values and note that they may change across PyTorch versions.
This flag is experimental and subject to change in future releases.
str, optional) —
The mode to use in torch.compile. If set to any value, torch_compile will be set to True.
Refer to the PyTorch doc for possible values and note that they may change across PyTorch versions.
This flag is experimental and subject to change in future releases.
bool, optional) —
Whether or not the accelerator should split the batches yielded by the dataloaders across the devices
during distributed training. If
set to True, the actual batch size used will be the same on any kind of distributed processes, but it
must be a
round multiple of the number of processes you are using (such as GPUs).
bool, optional) —
Whether or not to compute the number of tokens per second per device for training speed metrics.
This will iterate over the entire training dataloader once beforehand,
and will slow down the entire process.
bool, optional) —
Whether or not to track the number of input tokens seen throughout training.
May be slower in distributed training as gather operations must be called.
Optional[float]) —
If not None, this will activate NEFTune noise embeddings. This can drastically improve model performance
for instruction fine-tuning. Check out the original paper and the
original code. Support transformers PreTrainedModel and also
PeftModel from peft. bool, optional, defaults to False) —
Whether to use a sortish sampler or not. Only possible if the underlying datasets are Seq2SeqDataset
for now but will become generally available in the near future.
It sorts the inputs according to lengths in order to minimize the padding size, with a bit of randomness for the training set.
bool, optional, defaults to False) —
Whether to use generate to calculate generative metrics (ROUGE, BLEU). int, optional) —
The max_length to use on each evaluation loop when predict_with_generate=True. Will default to the
max_length value of the model configuration. int, optional) —
The num_beams to use on each evaluation loop when predict_with_generate=True. Will default to the
num_beams value of the model configuration. str or Path or GenerationConfig, optional) —
Allows to load a GenerationConfig from the from_pretrained method. This can be either:
./my_model_directory/.TrainingArguments is the subset of the arguments we use in our example scripts which relate to the training loop itself.
Using HfArgumentParser we can turn this class into argparse arguments that can be specified on the command line.
Serializes this instance while replace Enum by their values and GenerationConfig by dictionaries (for JSON
serialization support). It obfuscates the token values by removing their value.