TextClassifier¶
- class flash.text.classification.model.TextClassifier(num_classes=None, labels=None, backbone='prajjwal1/bert-medium', max_length=128, loss_fn=None, optimizer='Adam', lr_scheduler=None, metrics=None, learning_rate=None, multi_label=False, enable_ort=False)[source]¶
The
TextClassifier
is aTask
for classifying text. For more details, see Text Classification. TheTextClassifier
also supports multi-label classification withmulti_label=True
. For more details, see Multi-label Text Classification.- Parameters
num_classes¶ (
Optional
[int
]) – Number of classes to classify.backbone¶ (
str
) – A model to use to compute text features can be any BERT model from HuggingFace/transformersimage.max_length¶ (
int
) – The maximum length to pad / truncate sequences to.optimizer¶ (
TypeVar
(OPTIMIZER_TYPE
,str
,Callable
,Tuple
[str
,Dict
[str
,Any
]],None
)) – Optimizer to use for training.lr_scheduler¶ (
Optional
[TypeVar
(LR_SCHEDULER_TYPE
,str
,Callable
,Tuple
[str
,Dict
[str
,Any
]],Tuple
[str
,Dict
[str
,Any
],Dict
[str
,Any
]],None
)]) – The LR scheduler to use during training.metrics¶ (
Optional
[TypeVar
(METRICS_TYPE
,Metric
,Mapping
,Sequence
,None
)]) – Metrics to compute for training and evaluation. Can either be an metric from the torchmetrics package, a custom metric inherenting from torchmetrics.Metric, a callable function or a list/dict containing a combination of the aforementioned. In all cases, each metric needs to have the signature metric(preds,target) and return a single scalar tensor. Defaults totorchmetrics.Accuracy
.learning_rate¶ (
Optional
[float
]) – Learning rate to use for training, defaults to 1e-3multi_label¶ (
bool
) – Whether the targets are multi-label or not.enable_ort¶ (
bool
) – Enable Torch ONNX Runtime Optimization: https://onnxruntime.ai/docs/#onnx-runtime-for-training
- classmethod available_finetuning_strategies(cls)¶
Returns a list containing the keys of the available Finetuning Strategies.
- classmethod available_lr_schedulers(cls)¶
Returns a list containing the keys of the available LR schedulers.