Adding TF translation example (#12667)

* Adding TF translation example

* Fixes and style pass for TF translation example

* Remove unused postprocess_text copied from run_summarization

* Adding README

* Review fixes

* Move changes to model.config to after we've initialized the model

examples/tensorflow/translation/README.md

+<!---
+Copyright 2021 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+-->
+
+# Translation example
+
+This script shows an example of training a *translation* model with the 🤗 Transformers library.
+For straightforward use-cases you may be able to use these scripts without modification, although we have also
+included comments in the code to indicate areas that you may need to adapt to your own projects.
+
+### Multi-GPU and TPU usage
+
+By default, these scripts use a `MirroredStrategy` and will use multiple GPUs effectively if they are available. TPUs
+can also be used by passing the name of the TPU resource with the `--tpu` argument.
+
+### Example commands and caveats
+
+MBart and some T5 models require special handling.
+
+T5 models `t5-small`, `t5-base`, `t5-large`, `t5-3b` and `t5-11b` must use an additional argument: `--source_prefix "translate {source_lang} to {target_lang}"`. For example:
+
+```bash
+python run_translation.py \
+    --model_name_or_path t5-small \
+    --do_train \
+    --do_eval \
+    --source_lang en \
+    --target_lang ro \
+    --source_prefix "translate English to Romanian: " \
+    --dataset_name wmt16 \
+    --dataset_config_name ro-en \
+    --output_dir /tmp/tst-translation \
+    --per_device_train_batch_size=16 \
+    --per_device_eval_batch_size=16 \
+    --overwrite_output_dir
+```
+
+If you get a terrible BLEU score, make sure that you didn't forget to use the `--source_prefix` argument.
+
+For the aforementioned group of T5 models it's important to remember that if you switch to a different language pair, make sure to adjust the source and target values in all 3 language-specific command line argument: `--source_lang`, `--target_lang` and `--source_prefix`.
+
+MBart models require a different format for `--source_lang` and `--target_lang` values, e.g. instead of `en` it expects `en_XX`, for `ro` it expects `ro_RO`. The full MBart specification for language codes can be found [here](https://huggingface.co/facebook/mbart-large-cc25). For example:
+
+```bash
+python run_translation.py \
+    --model_name_or_path facebook/mbart-large-en-ro  \
+    --do_train \
+    --do_eval \
+    --dataset_name wmt16 \
+    --dataset_config_name ro-en \
+    --source_lang en_XX \
+    --target_lang ro_RO \
+    --output_dir /tmp/tst-translation \
+    --per_device_train_batch_size=16 \
+    --per_device_eval_batch_size=16 \
+    --overwrite_output_dir
+ ```

examples/tensorflow/translation/run_translation.py

+#!/usr/bin/env python
+# coding=utf-8
+# Copyright 2021 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+Fine-tuning the library models for translation.
+"""
+# You can also adapt this script on your own sequence to sequence task. Pointers for this are left as comments.
+
+import logging
+import os
+import sys
+from dataclasses import dataclass, field
+from functools import partial
+from typing import Optional
+
+import datasets
+import numpy as np
+import tensorflow as tf
+from datasets import load_dataset, load_metric
+from tqdm import tqdm
+
+import transformers
+from transformers import (
+    AutoConfig,
+    AutoTokenizer,
+    HfArgumentParser,
+    M2M100Tokenizer,
+    MBart50Tokenizer,
+    MBart50TokenizerFast,
+    MBartTokenizer,
+    MBartTokenizerFast,
+    TFAutoModelForSeq2SeqLM,
+    TFTrainingArguments,
+    create_optimizer,
+    set_seed,
+)
+from transformers.trainer_utils import get_last_checkpoint
+from transformers.utils import check_min_version
+from transformers.utils.versions import require_version
+
+
+# region Dependencies and constants
+# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
+check_min_version("4.9.0.dev0")
+
+require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/summarization/requirements.txt")
+
+logger = logging.getLogger(__name__)
+MULTILINGUAL_TOKENIZERS = [MBartTokenizer, MBartTokenizerFast, MBart50Tokenizer, MBart50TokenizerFast, M2M100Tokenizer]
+# endregion
+
+
+# region Arguments
+@dataclass
+class ModelArguments:
+    """
+    Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.
+    """
+
+    model_name_or_path: str = field(
+        metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"}
+    )
+    config_name: Optional[str] = field(
+        default=None, metadata={"help": "Pretrained config name or path if not the same as model_name"}
+    )
+    tokenizer_name: Optional[str] = field(
+        default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
+    )
+    cache_dir: Optional[str] = field(
+        default=None,
+        metadata={"help": "Where to store the pretrained models downloaded from huggingface.co"},
+    )
+    use_fast_tokenizer: bool = field(
+        default=True,
+        metadata={"help": "Whether to use one of the fast tokenizer (backed by the tokenizers library) or not."},
+    )
+    model_revision: str = field(
+        default="main",
+        metadata={"help": "The specific model version to use (can be a branch name, tag name or commit id)."},
+    )
+    use_auth_token: bool = field(
+        default=False,
+        metadata={
+            "help": "Will use the token generated when running `transformers-cli login` (necessary to use this script "
+            "with private models)."
+        },
+    )
+
+
+@dataclass
+class DataTrainingArguments:
+    """
+    Arguments pertaining to what data we are going to input our model for training and eval.
+    """
+
+    source_lang: str = field(default=None, metadata={"help": "Source language id for translation."})
+    target_lang: str = field(default=None, metadata={"help": "Target language id for translation."})
+    dataset_name: Optional[str] = field(
+        default=None, metadata={"help": "The name of the dataset to use (via the datasets library)."}
+    )
+    dataset_config_name: Optional[str] = field(
+        default=None, metadata={"help": "The configuration name of the dataset to use (via the datasets library)."}
+    )
+    train_file: Optional[str] = field(
+        default=None, metadata={"help": "The input training data file (a jsonlines or csv file)."}
+    )
+    validation_file: Optional[str] = field(
+        default=None,
+        metadata={
+            "help": "An optional input evaluation data file to evaluate the metrics (rouge) on "
+            "(a jsonlines or csv file)."
+        },
+    )
+    test_file: Optional[str] = field(
+        default=None,
+        metadata={
+            "help": "An optional input test data file to evaluate the metrics (rouge) on " "(a jsonlines or csv file)."
+        },
+    )
+    overwrite_cache: bool = field(
+        default=False, metadata={"help": "Overwrite the cached training and evaluation sets"}
+    )
+    preprocessing_num_workers: Optional[int] = field(
+        default=None,
+        metadata={"help": "The number of processes to use for the preprocessing."},
+    )
+    max_source_length: Optional[int] = field(
+        default=1024,
+        metadata={
+            "help": "The maximum total input sequence length after tokenization. Sequences longer "
+            "than this will be truncated, sequences shorter will be padded."
+        },
+    )
+    max_target_length: Optional[int] = field(
+        default=128,
+        metadata={
+            "help": "The maximum total sequence length for target text after tokenization. Sequences longer "
+            "than this will be truncated, sequences shorter will be padded."
+        },
+    )
+    val_max_target_length: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "The maximum total sequence length for validation target text after tokenization. Sequences longer "
+            "than this will be truncated, sequences shorter will be padded. Will default to `max_target_length`."
+            "This argument is also used to override the ``max_length`` param of ``model.generate``, which is used "
+            "during ``evaluate`` and ``predict``."
+        },
+    )
+    pad_to_max_length: bool = field(
+        default=False,
+        metadata={
+            "help": "Whether to pad all samples to model maximum sentence length. "
+            "If False, will pad the samples dynamically when batching to the maximum length in the batch. More "
+            "efficient on GPU but very bad for TPU."
+        },
+    )
+    max_train_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of training examples to this "
+            "value if set."
+        },
+    )
+    max_eval_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of evaluation examples to this "
+            "value if set."
+        },
+    )
+    max_predict_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of prediction examples to this "
+            "value if set."
+        },
+    )
+    num_beams: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "Number of beams to use for evaluation. This argument will be passed to ``model.generate``, "
+            "which is used during ``evaluate`` and ``predict``."
+        },
+    )
+    ignore_pad_token_for_loss: bool = field(
+        default=True,
+        metadata={
+            "help": "Whether to ignore the tokens corresponding to padded labels in the loss computation or not."
+        },
+    )
+    source_prefix: Optional[str] = field(
+        default=None, metadata={"help": "A prefix to add before every source text (useful for T5 models)."}
+    )
+
+    def __post_init__(self):
+        if self.dataset_name is None and self.train_file is None and self.validation_file is None:
+            raise ValueError("Need either a dataset name or a training/validation file.")
+        else:
+            if self.train_file is not None:
+                extension = self.train_file.split(".")[-1]
+                assert extension in ["csv", "json"], "`train_file` should be a csv or a json file."
+            if self.validation_file is not None:
+                extension = self.validation_file.split(".")[-1]
+                assert extension in ["csv", "json"], "`validation_file` should be a csv or a json file."
+        if self.val_max_target_length is None:
+            self.val_max_target_length = self.max_target_length
+
+
+# endregion
+
+# region Data generator
+def sample_generator(dataset, model, tokenizer, shuffle, pad_to_multiple_of=None):
+    if shuffle:
+        sample_ordering = np.random.permutation(len(dataset))
+    else:
+        sample_ordering = np.arange(len(dataset))
+    for sample_idx in sample_ordering:
+        example = dataset[int(sample_idx)]
+        # Handle dicts with proper padding and conversion to tensor.
+        example = tokenizer.pad(example, return_tensors="np", pad_to_multiple_of=pad_to_multiple_of)
+        example = {key: tf.convert_to_tensor(arr, dtype_hint=tf.int32) for key, arr in example.items()}
+        if model is not None and hasattr(model, "prepare_decoder_input_ids_from_labels"):
+            decoder_input_ids = model.prepare_decoder_input_ids_from_labels(
+                labels=tf.expand_dims(example["labels"], 0)
+            )
+            example["decoder_input_ids"] = tf.squeeze(decoder_input_ids, 0)
+        yield example, example["labels"]  # TF needs some kind of labels, even if we don't use them
+    return
+
+
+# endregion
+
+
+# region Helper functions
+def dataset_to_tf(dataset, model, tokenizer, total_batch_size, num_epochs, shuffle):
+    if dataset is None:
+        return None
+    train_generator = partial(sample_generator, dataset, model, tokenizer, shuffle=shuffle)
+    train_signature = {
+        feature: tf.TensorSpec(shape=(None,), dtype=tf.int32)
+        for feature in dataset.features
+        if feature != "special_tokens_mask"
+    }
+    if (
+        model is not None
+        and "decoder_input_ids" not in train_signature
+        and hasattr(model, "prepare_decoder_input_ids_from_labels")
+    ):
+        train_signature["decoder_input_ids"] = train_signature["labels"]
+    # This may need to be changed depending on your particular model or tokenizer!
+    padding_values = {
+        key: tf.convert_to_tensor(tokenizer.pad_token_id if tokenizer.pad_token_id is not None else 0, dtype=tf.int32)
+        for key in train_signature.keys()
+    }
+    padding_values["labels"] = tf.convert_to_tensor(-100, dtype=tf.int32)
+    train_signature["labels"] = train_signature["input_ids"]
+    train_signature = (train_signature, train_signature["labels"])
+    options = tf.data.Options()
+    options.experimental_distribute.auto_shard_policy = tf.data.experimental.AutoShardPolicy.OFF
+    tf_dataset = (
+        tf.data.Dataset.from_generator(train_generator, output_signature=train_signature)
+        .with_options(options)
+        .padded_batch(
+            batch_size=total_batch_size,
+            drop_remainder=True,
+            padding_values=(padding_values, np.array(-100, dtype=np.int32)),
+        )
+        .repeat(int(num_epochs))
+    )
+    return tf_dataset
+
+
+# endregion
+
+
+def main():
+    # region Argument parsing
+    # See all possible arguments in src/transformers/training_args.py
+    # or by passing the --help flag to this script.
+    # We now keep distinct sets of args, for a cleaner separation of concerns.
+
+    parser = HfArgumentParser((ModelArguments, DataTrainingArguments, TFTrainingArguments))
+    if len(sys.argv) == 2 and sys.argv[1].endswith(".json"):
+        # If we pass only one argument to the script and it's the path to a json file,
+        # let's parse it to get our arguments.
+        model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))
+    else:
+        model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+    # endregion
+
+    # region Logging
+    logging.basicConfig(
+        format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
+        datefmt="%m/%d/%Y %H:%M:%S",
+        handlers=[logging.StreamHandler(sys.stdout)],
+    )
+    logger.setLevel(logging.INFO)
+    datasets.utils.logging.set_verbosity(logging.INFO)
+    transformers.utils.logging.set_verbosity(logging.INFO)
+
+    # Log on each process the small summary:
+    logger.info(f"Training/evaluation parameters {training_args}")
+    # endregion
+
+    # region Detecting last checkpoint
+    last_checkpoint = None
+    if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir:
+        last_checkpoint = get_last_checkpoint(training_args.output_dir)
+        if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0:
+            raise ValueError(
+                f"Output directory ({training_args.output_dir}) already exists and is not empty. "
+                "Use --overwrite_output_dir to overcome."
+            )
+        elif last_checkpoint is not None and training_args.resume_from_checkpoint is None:
+            logger.info(
+                f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change "
+                "the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
+            )
+    # endregion
+
+    # Set seed before initializing model.
+    set_seed(training_args.seed)
+
+    # region Load datasets
+    # Get the datasets: you can either provide your own CSV/JSON training and evaluation files (see below)
+    # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/
+    # (the dataset will be downloaded automatically from the datasets Hub).
+    #
+    # For CSV/JSON files this script will use the first column for the full texts and the second column for the
+    # summaries (unless you specify column names for this with the `text_column` and `summary_column` arguments).
+    #
+    # In distributed training, the load_dataset function guarantee that only one local process can concurrently
+    # download the dataset.
+    if data_args.dataset_name is not None:
+        # Downloading and loading a dataset from the hub.
+        raw_datasets = load_dataset(
+            data_args.dataset_name, data_args.dataset_config_name, cache_dir=model_args.cache_dir
+        )
+    else:
+        data_files = {}
+        if data_args.train_file is not None:
+            data_files["train"] = data_args.train_file
+            extension = data_args.train_file.split(".")[-1]
+        if data_args.validation_file is not None:
+            data_files["validation"] = data_args.validation_file
+            extension = data_args.validation_file.split(".")[-1]
+        raw_datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
+    # See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at
+    # https://huggingface.co/docs/datasets/loading_datasets.html.
+    # endregion
+
+    # region Load model config and tokenizer
+    #
+    # Distributed training:
+    # The .from_pretrained methods guarantee that only one local process can concurrently
+    # download model & vocab.
+
+    config = AutoConfig.from_pretrained(
+        model_args.config_name if model_args.config_name else model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        revision=model_args.model_revision,
+        use_auth_token=True if model_args.use_auth_token else None,
+    )
+    tokenizer = AutoTokenizer.from_pretrained(
+        model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        use_fast=model_args.use_fast_tokenizer,
+        revision=model_args.model_revision,
+        use_auth_token=True if model_args.use_auth_token else None,
+    )
+
+    prefix = data_args.source_prefix if data_args.source_prefix is not None else ""
+    # endregion
+
+    # region Dataset preprocessing
+    # We need to tokenize inputs and targets.
+    if training_args.do_train:
+        column_names = raw_datasets["train"].column_names
+    elif training_args.do_eval:
+        column_names = raw_datasets["validation"].column_names
+    else:
+        logger.info("There is nothing to do. Please pass `do_train`, and/or `do_eval`.")
+        return
+
+    column_names = raw_datasets["train"].column_names
+
+    # For translation we set the codes of our source and target languages (only useful for mBART, the others will
+    # ignore those attributes).
+    if isinstance(tokenizer, tuple(MULTILINGUAL_TOKENIZERS)):
+        assert data_args.target_lang is not None and data_args.source_lang is not None, (
+            f"{tokenizer.__class__.__name__} is a multilingual tokenizer which requires --source_lang and "
+            "--target_lang arguments."
+        )
+        tokenizer.src_lang = data_args.source_lang
+        tokenizer.tgt_lang = data_args.target_lang
+        forced_bos_token_id = (
+            tokenizer.lang_code_to_id[data_args.forced_bos_token] if data_args.forced_bos_token is not None else None
+        )
+
+    # Get the language codes for input/target.
+    source_lang = data_args.source_lang.split("_")[0]
+    target_lang = data_args.target_lang.split("_")[0]
+
+    padding = "max_length" if data_args.pad_to_max_length else False
+
+    # Temporarily set max_target_length for training.
+    max_target_length = data_args.max_target_length
+    padding = "max_length" if data_args.pad_to_max_length else False
+
+    def preprocess_function(examples):
+        inputs = [ex[source_lang] for ex in examples["translation"]]
+        targets = [ex[target_lang] for ex in examples["translation"]]
+        inputs = [prefix + inp for inp in inputs]
+        model_inputs = tokenizer(inputs, max_length=data_args.max_source_length, padding=padding, truncation=True)
+
+        # Setup the tokenizer for targets
+        with tokenizer.as_target_tokenizer():
+            labels = tokenizer(targets, max_length=max_target_length, padding=padding, truncation=True)
+
+        # If we are padding here, replace all tokenizer.pad_token_id in the labels by -100 when we want to ignore
+        # padding in the loss.
+        if padding == "max_length" and data_args.ignore_pad_token_for_loss:
+            labels["input_ids"] = [
+                [(l if l != tokenizer.pad_token_id else -100) for l in label] for label in labels["input_ids"]
+            ]
+
+        model_inputs["labels"] = labels["input_ids"]
+        return model_inputs
+
+    if training_args.do_train:
+        if "train" not in raw_datasets:
+            raise ValueError("--do_train requires a train dataset")
+        train_dataset = raw_datasets["train"]
+        if data_args.max_train_samples is not None:
+            train_dataset = train_dataset.select(range(data_args.max_train_samples))
+        with training_args.main_process_first(desc="train dataset map pre-processing"):
+            train_dataset = train_dataset.map(
+                preprocess_function,
+                batched=True,
+                num_proc=data_args.preprocessing_num_workers,
+                remove_columns=column_names,
+                load_from_cache_file=not data_args.overwrite_cache,
+                desc="Running tokenizer on train dataset",
+            )
+    else:
+        train_dataset = None
+
+    if training_args.do_eval:
+        max_target_length = data_args.val_max_target_length
+        if "validation" not in raw_datasets:
+            raise ValueError("--do_eval requires a validation dataset")
+        eval_dataset = raw_datasets["validation"]
+        if data_args.max_eval_samples is not None:
+            eval_dataset = eval_dataset.select(range(data_args.max_eval_samples))
+        with training_args.main_process_first(desc="validation dataset map pre-processing"):
+            eval_dataset = eval_dataset.map(
+                preprocess_function,
+                batched=True,
+                num_proc=data_args.preprocessing_num_workers,
+                remove_columns=column_names,
+                load_from_cache_file=not data_args.overwrite_cache,
+                desc="Running tokenizer on validation dataset",
+            )
+    else:
+        eval_dataset = None
+    # endregion
+
+    with training_args.strategy.scope():
+        # region Prepare model
+        model = TFAutoModelForSeq2SeqLM.from_pretrained(
+            model_args.model_name_or_path,
+            config=config,
+            cache_dir=model_args.cache_dir,
+            revision=model_args.model_revision,
+            use_auth_token=True if model_args.use_auth_token else None,
+        )
+
+        model.resize_token_embeddings(len(tokenizer))
+        if isinstance(tokenizer, tuple(MULTILINGUAL_TOKENIZERS)):
+            model.config.forced_bos_token_id = forced_bos_token_id
+        # endregion
+
+        # region Set decoder_start_token_id
+        if model.config.decoder_start_token_id is None and isinstance(tokenizer, (MBartTokenizer, MBartTokenizerFast)):
+            assert (
+                data_args.target_lang is not None and data_args.source_lang is not None
+            ), "mBart requires --target_lang and --source_lang"
+            if isinstance(tokenizer, MBartTokenizer):
+                model.config.decoder_start_token_id = tokenizer.lang_code_to_id[data_args.target_lang]
+            else:
+                model.config.decoder_start_token_id = tokenizer.convert_tokens_to_ids(data_args.target_lang)
+
+        if model.config.decoder_start_token_id is None:
+            raise ValueError("Make sure that `config.decoder_start_token_id` is correctly defined")
+        # endregion
+
+        # region Prepare TF Dataset objects
+        num_replicas = training_args.strategy.num_replicas_in_sync
+        total_train_batch_size = training_args.per_device_train_batch_size * num_replicas
+        total_eval_batch_size = training_args.per_device_eval_batch_size * num_replicas
+        tf_train_dataset = dataset_to_tf(
+            train_dataset,
+            model,
+            tokenizer,
+            total_batch_size=total_train_batch_size,
+            num_epochs=training_args.num_train_epochs,
+            shuffle=True,
+        )
+        tf_eval_dataset = dataset_to_tf(
+            eval_dataset,
+            model,
+            tokenizer,
+            total_eval_batch_size,
+            num_epochs=1,
+            shuffle=False,
+        )
+        # endregion
+
+        # region Optimizer, loss and LR scheduling
+        # Scheduler and math around the number of training steps.
+        num_update_steps_per_epoch = len(train_dataset) // training_args.per_device_train_batch_size
+        num_train_steps = training_args.num_train_epochs * num_update_steps_per_epoch
+        optimizer, lr_schedule = create_optimizer(
+            init_lr=training_args.learning_rate,
+            num_train_steps=num_train_steps,
+            num_warmup_steps=training_args.warmup_steps,
+        )
+
+        def masked_sparse_categorical_crossentropy(y_true, y_pred):
+            # We clip the negative labels to 0 to avoid NaNs appearing in the output and
+            # fouling up everything that comes afterwards. The loss values corresponding to clipped values
+            # will be masked later anyway, but even masked NaNs seem to cause overflows for some reason.
+            # 1e6 is chosen as a reasonable upper bound for the number of token indices - in the unlikely
+            # event that you have more than 1 million tokens in your vocabulary, consider increasing this value.
+            # More pragmatically, consider redesigning your tokenizer.
+            losses = tf.keras.losses.sparse_categorical_crossentropy(
+                tf.clip_by_value(y_true, 0, int(1e6)), y_pred, from_logits=True
+            )
+            # Compute the per-sample loss only over the unmasked tokens
+            losses = tf.ragged.boolean_mask(losses, y_true != -100)
+            losses = tf.reduce_mean(losses, axis=-1)
+            return losses
+
+        # endregion
+
+        # region Metric and postprocessing
+        metric = load_metric("sacrebleu")
+
+        def postprocess_text(preds, labels):
+            preds = [pred.strip() for pred in preds]
+            labels = [[label.strip()] for label in labels]
+
+            return preds, labels
+
+        # endregion
+
+        # region Training
+        model.compile(loss={"logits": masked_sparse_categorical_crossentropy}, optimizer=optimizer)
+
+        if training_args.do_train:
+            logger.info("***** Running training *****")
+            logger.info(f"  Num examples = {len(train_dataset)}")
+            logger.info(f"  Num Epochs = {training_args.num_train_epochs}")
+            logger.info(f"  Instantaneous batch size per device = {training_args.per_device_train_batch_size}")
+            logger.info(f"  Total train batch size = {total_train_batch_size}")
+            logger.info(f"  Total optimization steps = {num_train_steps}")
+
+            model.fit(
+                tf_train_dataset,
+                epochs=int(training_args.num_train_epochs),
+                steps_per_epoch=num_update_steps_per_epoch,
+            )
+        # endregion
+
+        # region Validation
+        if data_args.val_max_target_length is None:
+            data_args.val_max_target_length = data_args.max_target_length
+
+        gen_kwargs = {
+            "max_length": data_args.val_max_target_length,
+            "num_beams": data_args.num_beams,
+        }
+        if training_args.do_eval:
+            logger.info("Evaluation...")
+            for batch, labels in tqdm(
+                tf_eval_dataset, total=len(eval_dataset) // training_args.per_device_eval_batch_size
+            ):
+                batch.update(gen_kwargs)
+                generated_tokens = model.generate(**batch)
+                if isinstance(generated_tokens, tuple):
+                    generated_tokens = generated_tokens[0]
+                decoded_preds = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
+                labels = np.where(labels != -100, labels, tokenizer.pad_token_id)
+                decoded_labels = tokenizer.batch_decode(labels, skip_special_tokens=True)
+                decoded_preds, decoded_labels = postprocess_text(decoded_preds, decoded_labels)
+
+                metric.add_batch(predictions=decoded_preds, references=decoded_labels)
+            eval_metric = metric.compute()
+            logger.info({"bleu": eval_metric["score"]})
+        # endregion
+
+        if training_args.output_dir is not None:
+            model.save_pretrained(training_args.output_dir)
+
+
+if __name__ == "__main__":
+    main()