BuilderConfig for Masakhaner

Represents a sparse fine-tuning of a pre-trained base model. Contains two sets of tensors, "diffs", the difference tensors for sparsely fine-

Data collator that will dynamically pad the inputs received, as well as the labels. Args: tokenizer (:class:`~transformers.PreTraine

based on allennlp.training.metrics.AttachmentScores Computes labeled and unlabeled attachment scores for a dependency parse. Note that the in

Arguments pertaining to what data we are going to input our model for training and eval.

Arguments pertaining to what data we are going to input our model for training and eval.

Arguments pertaining to what data we are going to input our model for training and eval. Using `HfArgumentParser` we can turn this class

Arguments pertaining to what data we are going to input our model for training and eval.

Arguments pertaining to what data we are going to input our model for training and eval.

Masakhaner dataset.

Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.

Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.

Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.

Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.

Arguments pertaining to which model/config/tokenizer we are going to fine-tune, or train from scratch.

Credit: G. Glavaš & I. Vulić Based on paper "Is Supervised Syntactic Parsing Beneficial for Language Understanding? An Empirical Investigatio

Extends TrainingArguments for Universal Dependencies (UD) dependency parsing. TrainingArguments is the subset of the arguments we use in our

Superclass for Trainers that learn sparse fine-tunings. Keeps track of original model parameters so that difference vectors can be calculated