Add eval scripts

evaluation/paws.yaml

@@ -0,0 +1,55 @@
+name: BERTIN PAWS-X es
+project: bertin-eval
+enitity: versae
+program: run_glue.py
+command:
+  - ${env}
+  - ${interpreter}
+  - ${program}
+  - ${args}
+method: grid
+metric:
+  name: eval/accuracy
+  goal: maximize
+parameters:
+  model_name_or_path:
+    values:
+    - bertin-project/bertin-base-gaussian-exp-512seqlen
+    - bertin-project/bertin-base-random-exp-512seqlen
+    - bertin-project/bertin-base-gaussian
+    - bertin-project/bertin-base-stepwise
+    - bertin-project/bertin-base-random
+    - bertin-project/bertin-roberta-base-spanish
+    - flax-community/bertin-roberta-large-spanish
+    - BSC-TeMU/roberta-base-bne
+    - dccuchile/bert-base-spanish-wwm-cased
+    - bert-base-multilingual-cased
+  num_train_epochs:
+    values: [5]
+  task_name:
+    value: paws-x
+  dataset_name:
+    value: paws-x
+  dataset_config_name:
+    value: es
+  output_dir:
+    value: ./outputs
+  overwrite_output_dir:
+    value: true
+  resume_from_checkpoint:
+    value: false
+  max_seq_length:
+    value: 512
+  pad_to_max_length:
+    value: true
+  per_device_train_batch_size:
+    value: 16
+  per_device_eval_batch_size:
+    value: 16
+  save_total_limit:
+    value: 1
+  do_train:
+    value: true
+  do_eval:
+    value: true
+

evaluation/run_glue.py

@@ -0,0 +1,576 @@
+#!/usr/bin/env python
+# coding=utf-8
+# Copyright 2020 The HuggingFace Inc. 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.
+""" Finetuning the library models for sequence classification on GLUE."""
+# You can also adapt this script on your own text classification task. Pointers for this are left as comments.
+
+import logging
+import os
+import random
+import sys
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Optional
+
+import datasets
+import numpy as np
+from datasets import load_dataset, load_metric
+
+import transformers
+from transformers import (
+    AutoConfig,
+    AutoModelForSequenceClassification,
+    AutoTokenizer,
+    DataCollatorWithPadding,
+    EvalPrediction,
+    HfArgumentParser,
+    PretrainedConfig,
+    Trainer,
+    TrainingArguments,
+    default_data_collator,
+    set_seed,
+)
+from transformers.trainer_utils import get_last_checkpoint
+from transformers.utils import check_min_version
+from transformers.utils.versions import require_version
+
+
+# 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/text-classification/requirements.txt")
+
+task_to_keys = {
+    "cola": ("sentence", None),
+    "mnli": ("premise", "hypothesis"),
+    "xnli": ("premise", "hypothesis"),
+    "mrpc": ("sentence1", "sentence2"),
+    "qnli": ("question", "sentence"),
+    "qqp": ("question1", "question2"),
+    "rte": ("sentence1", "sentence2"),
+    "sst2": ("sentence", None),
+    "stsb": ("sentence1", "sentence2"),
+    "wnli": ("sentence1", "sentence2"),
+    "paws-x": ("sentence1", "sentence2"),
+}
+task_to_metrics = {
+    "paws-x": "accuracy",
+    "xnli": "accuracy",
+}
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class DataTrainingArguments:
+    """
+    Arguments pertaining to what data we are going to input our model for training and eval.
+
+    Using `HfArgumentParser` we can turn this class
+    into argparse arguments to be able to specify them on
+    the command line.
+    """
+
+    task_name: Optional[str] = field(
+        default=None,
+        metadata={"help": "The name of the task to train on: " + ", ".join(task_to_keys.keys())},
+    )
+    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)."}
+    )
+    max_seq_length: int = field(
+        default=128,
+        metadata={
+            "help": "The maximum total input sequence length after tokenization. Sequences longer "
+            "than this will be truncated, sequences shorter will be padded."
+        },
+    )
+    overwrite_cache: bool = field(
+        default=False, metadata={"help": "Overwrite the cached preprocessed datasets or not."}
+    )
+    pad_to_max_length: bool = field(
+        default=True,
+        metadata={
+            "help": "Whether to pad all samples to `max_seq_length`. "
+            "If False, will pad the samples dynamically when batching to the maximum length in the batch."
+        },
+    )
+    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."
+        },
+    )
+    train_file: Optional[str] = field(
+        default=None, metadata={"help": "A csv or a json file containing the training data."}
+    )
+    validation_file: Optional[str] = field(
+        default=None, metadata={"help": "A csv or a json file containing the validation data."}
+    )
+    test_file: Optional[str] = field(default=None, metadata={"help": "A csv or a json file containing the test data."})
+
+    def __post_init__(self):
+        if self.task_name is not None:
+            self.task_name = self.task_name.lower()
+            if self.task_name not in task_to_keys.keys():
+                raise ValueError("Unknown task, you should pick one in " + ",".join(task_to_keys.keys()))
+        elif self.dataset_name is not None:
+            pass
+        elif self.train_file is None or self.validation_file is None:
+            raise ValueError("Need either a GLUE task, a training/validation file or a dataset name.")
+        else:
+            train_extension = self.train_file.split(".")[-1]
+            assert train_extension in ["csv", "json"], "`train_file` should be a csv or a json file."
+            validation_extension = self.validation_file.split(".")[-1]
+            assert (
+                validation_extension == train_extension
+            ), "`validation_file` should have the same extension (csv or json) as `train_file`."
+
+
+@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 do you want 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)."
+        },
+    )
+
+
+def main():
+    # 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, TrainingArguments))
+    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()
+
+    # Setup logging
+    logging.basicConfig(
+        format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
+        datefmt="%m/%d/%Y %H:%M:%S",
+        handlers=[logging.StreamHandler(sys.stdout)],
+    )
+
+    log_level = training_args.get_process_log_level()
+    logger.setLevel(log_level)
+    datasets.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.enable_default_handler()
+    transformers.utils.logging.enable_explicit_format()
+
+    # Log on each process the small summary:
+    logger.warning(
+        f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
+        + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
+    )
+    logger.info(f"Training/evaluation parameters {training_args}")
+
+    # Detecting last checkpoint.
+    last_checkpoint = None
+    run_name = f"{model_args.model_name_or_path}-{np.random.randint(1000):04d}"
+    training_args.output_dir = str(Path(training_args.output_dir) / run_name)
+    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."
+            )
+
+    # Set seed before initializing model.
+    set_seed(training_args.seed)
+
+    # Get the datasets: you can either provide your own CSV/JSON training and evaluation files (see below)
+    # or specify a GLUE benchmark task (the dataset will be downloaded automatically from the datasets Hub).
+    #
+    # For CSV/JSON files, this script will use as labels the column called 'label' and as pair of sentences the
+    # sentences in columns called 'sentence1' and 'sentence2' if such column exists or the first two columns not named
+    # label if at least two columns are provided.
+    #
+    # If the CSVs/JSONs contain only one non-label column, the script does single sentence classification on this
+    # single column. You can easily tweak this behavior (see below)
+    #
+    # 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
+        )
+    elif data_args.task_name is not None:
+        # Downloading and loading a dataset from the hub.
+        raw_datasets = load_dataset("glue", data_args.task_name, cache_dir=model_args.cache_dir)
+    else:
+        # Loading a dataset from your local files.
+        # CSV/JSON training and evaluation files are needed.
+        data_files = {"train": data_args.train_file, "validation": data_args.validation_file}
+
+        # Get the test dataset: you can provide your own CSV/JSON test file (see below)
+        # when you use `do_predict` without specifying a GLUE benchmark task.
+        if training_args.do_predict:
+            if data_args.test_file is not None:
+                train_extension = data_args.train_file.split(".")[-1]
+                test_extension = data_args.test_file.split(".")[-1]
+                assert (
+                    test_extension == train_extension
+                ), "`test_file` should have the same extension (csv or json) as `train_file`."
+                data_files["test"] = data_args.test_file
+            else:
+                raise ValueError("Need either a GLUE task or a test file for `do_predict`.")
+
+        for key in data_files.keys():
+            logger.info(f"load a local file for {key}: {data_files[key]}")
+
+        if data_args.train_file.endswith(".csv"):
+            # Loading a dataset from local csv files
+            raw_datasets = load_dataset("csv", data_files=data_files, cache_dir=model_args.cache_dir)
+        else:
+            # Loading a dataset from local json files
+            raw_datasets = load_dataset("json", data_files=data_files, cache_dir=model_args.cache_dir)
+    # See more about loading any type of standard or custom dataset at
+    # https://huggingface.co/docs/datasets/loading_datasets.html.
+
+    # Labels
+    if data_args.task_name is not None:
+        is_regression = data_args.task_name == "stsb"
+        if not is_regression:
+            label_list = raw_datasets["train"].features["label"].names
+            num_labels = len(label_list)
+        else:
+            num_labels = 1
+    else:
+        # Trying to have good defaults here, don't hesitate to tweak to your needs.
+        is_regression = raw_datasets["train"].features["label"].dtype in ["float32", "float64"]
+        if is_regression:
+            num_labels = 1
+        else:
+            # A useful fast method:
+            # https://huggingface.co/docs/datasets/package_reference/main_classes.html#datasets.Dataset.unique
+            label_list = raw_datasets["train"].unique("label")
+            label_list.sort()  # Let's sort it for determinism
+            num_labels = len(label_list)
+
+    # Load pretrained model and tokenizer
+    #
+    # In 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,
+        num_labels=num_labels,
+        finetuning_task=data_args.task_name,
+        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,
+    )
+    model = AutoModelForSequenceClassification.from_pretrained(
+        model_args.model_name_or_path,
+        from_tf=bool(".ckpt" in 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,
+    )
+    tokenizer.model_max_length = 512
+
+    # Preprocessing the raw_datasets
+    if data_args.task_name is not None:
+        sentence1_key, sentence2_key = task_to_keys[data_args.task_name]
+    else:
+        # Again, we try to have some nice defaults but don't hesitate to tweak to your use case.
+        non_label_column_names = [name for name in raw_datasets["train"].column_names if name != "label"]
+        if "sentence1" in non_label_column_names and "sentence2" in non_label_column_names:
+            sentence1_key, sentence2_key = "sentence1", "sentence2"
+        else:
+            if len(non_label_column_names) >= 2:
+                sentence1_key, sentence2_key = non_label_column_names[:2]
+            else:
+                sentence1_key, sentence2_key = non_label_column_names[0], None
+
+    # Padding strategy
+    if data_args.pad_to_max_length:
+        padding = "max_length"
+    else:
+        # We will pad later, dynamically at batch creation, to the max sequence length in each batch
+        padding = False
+
+    # Some models have set the order of the labels to use, so let's make sure we do use it.
+    label_to_id = None
+    if (
+        model.config.label2id != PretrainedConfig(num_labels=num_labels).label2id
+        and data_args.task_name is not None
+        and not is_regression
+    ):
+        # Some have all caps in their config, some don't.
+        label_name_to_id = {k.lower(): v for k, v in model.config.label2id.items()}
+        if list(sorted(label_name_to_id.keys())) == list(sorted(label_list)):
+            label_to_id = {i: int(label_name_to_id[label_list[i]]) for i in range(num_labels)}
+        else:
+            logger.warning(
+                "Your model seems to have been trained with labels, but they don't match the dataset: ",
+                f"model labels: {list(sorted(label_name_to_id.keys()))}, dataset labels: {list(sorted(label_list))}."
+                "\nIgnoring the model labels as a result.",
+            )
+    elif data_args.task_name is None and not is_regression:
+        label_to_id = {v: i for i, v in enumerate(label_list)}
+
+    if label_to_id is not None:
+        model.config.label2id = label_to_id
+        model.config.id2label = {id: label for label, id in config.label2id.items()}
+
+    if data_args.max_seq_length > tokenizer.model_max_length:
+        logger.warning(
+            f"The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the"
+            f"model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}."
+        )
+    max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length)
+
+    def preprocess_function(examples):
+        # Tokenize the texts
+        args = (
+            (examples[sentence1_key],) if sentence2_key is None else (examples[sentence1_key], examples[sentence2_key])
+        )
+        result = tokenizer(*args, padding=padding, max_length=max_seq_length, truncation=True)
+
+        # Map labels to IDs (not necessary for GLUE tasks)
+        if label_to_id is not None and "label" in examples:
+            result["label"] = [(label_to_id[l] if l != -1 else -1) for l in examples["label"]]
+        return result
+
+    with training_args.main_process_first(desc="dataset map pre-processing"):
+        raw_datasets = raw_datasets.map(
+            preprocess_function,
+            batched=True,
+            load_from_cache_file=not data_args.overwrite_cache,
+            desc="Running tokenizer on dataset",
+        )
+    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))
+
+    if training_args.do_eval:
+        if "validation" not in raw_datasets and "validation_matched" not in raw_datasets:
+            raise ValueError("--do_eval requires a validation dataset")
+        eval_dataset = raw_datasets["validation_matched" if data_args.task_name == "mnli" else "validation"]
+        if data_args.max_eval_samples is not None:
+            eval_dataset = eval_dataset.select(range(data_args.max_eval_samples))
+
+    if training_args.do_predict or data_args.task_name is not None or data_args.test_file is not None:
+        if "test" not in raw_datasets and "test_matched" not in raw_datasets:
+            raise ValueError("--do_predict requires a test dataset")
+        predict_dataset = raw_datasets["test_matched" if data_args.task_name == "mnli" else "test"]
+        if data_args.max_predict_samples is not None:
+            predict_dataset = predict_dataset.select(range(data_args.max_predict_samples))
+
+    # Log a few random samples from the training set:
+    if training_args.do_train:
+        for index in random.sample(range(len(train_dataset)), 3):
+            logger.info(f"Sample {index} of the training set: {train_dataset[index]}.")
+
+    # Get the metric function
+    if data_args.task_name in task_to_metrics:
+        metric = load_metric(task_to_metrics[data_args.task_name])
+    elif data_args.task_name is not None:
+        metric = load_metric("glue", data_args.task_name)
+    else:
+        metric = load_metric("accuracy")
+
+    # You can define your custom compute_metrics function. It takes an `EvalPrediction` object (a namedtuple with a
+    # predictions and label_ids field) and has to return a dictionary string to float.
+    def compute_metrics(p: EvalPrediction):
+        preds = p.predictions[0] if isinstance(p.predictions, tuple) else p.predictions
+        preds = np.squeeze(preds) if is_regression else np.argmax(preds, axis=1)
+        if data_args.task_name is not None:
+            result = metric.compute(predictions=preds, references=p.label_ids)
+            if len(result) > 1:
+                result["combined_score"] = np.mean(list(result.values())).item()
+            return result
+        elif is_regression:
+            return {"mse": ((preds - p.label_ids) ** 2).mean().item()}
+        else:
+            return {"accuracy": (preds == p.label_ids).astype(np.float32).mean().item()}
+
+    # Data collator will default to DataCollatorWithPadding, so we change it if we already did the padding.
+    if data_args.pad_to_max_length:
+        data_collator = default_data_collator
+    elif training_args.fp16:
+        data_collator = DataCollatorWithPadding(tokenizer, pad_to_multiple_of=8)
+    else:
+        data_collator = None
+
+    training_args.run_name = run_name
+    # Initialize our Trainer
+    trainer = Trainer(
+        model=model,
+        args=training_args,
+        train_dataset=train_dataset if training_args.do_train else None,
+        eval_dataset=eval_dataset if training_args.do_eval else None,
+        compute_metrics=compute_metrics,
+        tokenizer=tokenizer,
+        data_collator=data_collator,
+    )
+
+    # Training
+    if training_args.do_train:
+        checkpoint = None
+        if training_args.resume_from_checkpoint is not None:
+            checkpoint = training_args.resume_from_checkpoint
+        elif last_checkpoint is not None:
+            checkpoint = last_checkpoint
+        train_result = trainer.train(resume_from_checkpoint=checkpoint)
+        metrics = train_result.metrics
+        max_train_samples = (
+            data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset)
+        )
+        metrics["train_samples"] = min(max_train_samples, len(train_dataset))
+
+        trainer.save_model()  # Saves the tokenizer too for easy upload
+
+        trainer.log_metrics("train", metrics)
+        trainer.save_metrics("train", metrics)
+        trainer.save_state()
+
+    # Evaluation
+    if training_args.do_eval:
+        logger.info("*** Evaluate ***")
+
+        # Loop to handle MNLI double evaluation (matched, mis-matched)
+        tasks = [data_args.task_name]
+        eval_datasets = [eval_dataset]
+        if data_args.task_name == "mnli":
+            tasks.append("mnli-mm")
+            eval_datasets.append(raw_datasets["validation_mismatched"])
+
+        for eval_dataset, task in zip(eval_datasets, tasks):
+            metrics = trainer.evaluate(eval_dataset=eval_dataset)
+
+            max_eval_samples = (
+                data_args.max_eval_samples if data_args.max_eval_samples is not None else len(eval_dataset)
+            )
+            metrics["eval_samples"] = min(max_eval_samples, len(eval_dataset))
+
+            trainer.log_metrics("eval", metrics)
+            trainer.save_metrics("eval", metrics)
+
+    if training_args.do_predict:
+        logger.info("*** Predict ***")
+
+        # Loop to handle MNLI double evaluation (matched, mis-matched)
+        tasks = [data_args.task_name]
+        predict_datasets = [predict_dataset]
+        if data_args.task_name == "mnli":
+            tasks.append("mnli-mm")
+            predict_datasets.append(raw_datasets["test_mismatched"])
+
+        for predict_dataset, task in zip(predict_datasets, tasks):
+            # Removing the `label` columns because it contains -1 and Trainer won't like that.
+            predict_dataset = predict_dataset.remove_columns("label")
+            predictions = trainer.predict(predict_dataset, metric_key_prefix="predict").predictions
+            predictions = np.squeeze(predictions) if is_regression else np.argmax(predictions, axis=1)
+
+            output_predict_file = os.path.join(training_args.output_dir, f"predict_results_{task}.txt")
+            if trainer.is_world_process_zero():
+                with open(output_predict_file, "w") as writer:
+                    logger.info(f"***** Predict results {task} *****")
+                    writer.write("index\tprediction\n")
+                    for index, item in enumerate(predictions):
+                        if is_regression:
+                            writer.write(f"{index}\t{item:3.3f}\n")
+                        else:
+                            item = label_list[item]
+                            writer.write(f"{index}\t{item}\n")
+
+    if training_args.push_to_hub:
+        kwargs = {"finetuned_from": model_args.model_name_or_path, "tasks": "text-classification"}
+        if data_args.task_name is not None:
+            kwargs["language"] = "en"
+            kwargs["dataset_tags"] = "glue"
+            kwargs["dataset_args"] = data_args.task_name
+            kwargs["dataset"] = f"GLUE {data_args.task_name.upper()}"
+
+        trainer.push_to_hub(**kwargs)
+
+
+def _mp_fn(index):
+    # For xla_spawn (TPUs)
+    main()
+
+
+if __name__ == "__main__":
+    main()

evaluation/run_ner.py

@@ -0,0 +1,562 @@
+#!/usr/bin/env python
+# coding=utf-8
+# Copyright 2020 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 token classification.
+"""
+# You can also adapt this script on your own token classification task and datasets. Pointers for this are left as
+# comments.
+
+import logging
+import os
+import sys
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Optional
+
+import datasets
+import numpy as np
+from datasets import ClassLabel, load_dataset, load_metric
+
+import transformers
+from transformers import (
+    AutoConfig,
+    AutoModelForTokenClassification,
+    AutoTokenizer,
+    DataCollatorForTokenClassification,
+    HfArgumentParser,
+    PreTrainedTokenizerFast,
+    Trainer,
+    TrainingArguments,
+    set_seed,
+)
+from transformers.trainer_utils import get_last_checkpoint
+from transformers.utils import check_min_version
+from transformers.utils.versions import require_version
+
+
+# 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/token-classification/requirements.txt")
+
+logger = logging.getLogger(__name__)
+
+
+@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 do you want to store the pretrained models downloaded from huggingface.co"},
+    )
+    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.
+    """
+
+    task_name: Optional[str] = field(default="ner", metadata={"help": "The name of the task (ner, pos...)."})
+    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 csv or JSON file)."}
+    )
+    validation_file: Optional[str] = field(
+        default=None,
+        metadata={"help": "An optional input evaluation data file to evaluate on (a csv or JSON file)."},
+    )
+    test_file: Optional[str] = field(
+        default=None,
+        metadata={"help": "An optional input test data file to predict on (a csv or JSON file)."},
+    )
+    text_column_name: Optional[str] = field(
+        default=None, metadata={"help": "The column name of text to input in the file (a csv or JSON file)."}
+    )
+    label_column_name: Optional[str] = field(
+        default=None, metadata={"help": "The column name of label to input in the file (a csv or JSON 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."},
+    )
+    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."
+        },
+    )
+    label_all_tokens: bool = field(
+        default=False,
+        metadata={
+            "help": "Whether to put the label for one word on all tokens of generated by that word or just on the "
+            "one (in which case the other tokens will have a padding index)."
+        },
+    )
+    return_entity_level_metrics: bool = field(
+        default=False,
+        metadata={"help": "Whether to return all the entity levels during evaluation or just the overall ones."},
+    )
+
+    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."
+        self.task_name = self.task_name.lower()
+
+
+def main():
+    # 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, TrainingArguments))
+    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()
+
+    # Setup logging
+    logging.basicConfig(
+        format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
+        datefmt="%m/%d/%Y %H:%M:%S",
+        handlers=[logging.StreamHandler(sys.stdout)],
+    )
+
+    log_level = training_args.get_process_log_level()
+    logger.setLevel(log_level)
+    datasets.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.enable_default_handler()
+    transformers.utils.logging.enable_explicit_format()
+
+    # Log on each process the small summary:
+    logger.warning(
+        f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
+        + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
+    )
+    logger.info(f"Training/evaluation parameters {training_args}")
+
+    # Detecting last checkpoint.
+    last_checkpoint = None
+    run_name = f"{model_args.model_name_or_path}-{np.random.randint(1000):04d}"
+    training_args.output_dir = str(Path(training_args.output_dir) / run_name)
+    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."
+            )
+
+    # Set seed before initializing model.
+    set_seed(training_args.seed)
+
+    # Get the datasets: you can either provide your own CSV/JSON/TXT 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 column called 'text' or the first column if no column called
+    # 'text' is found. You can easily tweak this behavior (see below).
+    #
+    # 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
+        if data_args.validation_file is not None:
+            data_files["validation"] = data_args.validation_file
+        if data_args.test_file is not None:
+            data_files["test"] = data_args.test_file
+        extension = data_args.train_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.
+
+    if training_args.do_train:
+        column_names = raw_datasets["train"].column_names
+        features = raw_datasets["train"].features
+    else:
+        column_names = raw_datasets["validation"].column_names
+        features = raw_datasets["validation"].features
+
+    if data_args.text_column_name is not None:
+        text_column_name = data_args.text_column_name
+    elif "tokens" in column_names:
+        text_column_name = "tokens"
+    else:
+        text_column_name = column_names[0]
+
+    if data_args.label_column_name is not None:
+        label_column_name = data_args.label_column_name
+    elif f"{data_args.task_name}_tags" in column_names:
+        label_column_name = f"{data_args.task_name}_tags"
+    else:
+        label_column_name = column_names[1]
+
+    # In the event the labels are not a `Sequence[ClassLabel]`, we will need to go through the dataset to get the
+    # unique labels.
+    def get_label_list(labels):
+        unique_labels = set()
+        for label in labels:
+            unique_labels = unique_labels | set(label)
+        label_list = list(unique_labels)
+        label_list.sort()
+        return label_list
+
+    if isinstance(features[label_column_name].feature, ClassLabel):
+        label_list = features[label_column_name].feature.names
+        # No need to convert the labels since they are already ints.
+        label_to_id = {i: i for i in range(len(label_list))}
+    else:
+        label_list = get_label_list(raw_datasets["train"][label_column_name])
+        label_to_id = {l: i for i, l in enumerate(label_list)}
+    num_labels = len(label_list)
+
+    # Load pretrained model 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,
+        num_labels=num_labels,
+        label2id=label_to_id,
+        id2label={i: l for l, i in label_to_id.items()},
+        finetuning_task=data_args.task_name,
+        cache_dir=model_args.cache_dir,
+        revision=model_args.model_revision,
+        use_auth_token=True if model_args.use_auth_token else None,
+    )
+
+    tokenizer_name_or_path = model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path
+    if config.model_type in {"gpt2", "roberta"}:
+        tokenizer = AutoTokenizer.from_pretrained(
+            tokenizer_name_or_path,
+            cache_dir=model_args.cache_dir,
+            use_fast=True,
+            revision=model_args.model_revision,
+            use_auth_token=True if model_args.use_auth_token else None,
+            add_prefix_space=True,
+        )
+    else:
+        tokenizer = AutoTokenizer.from_pretrained(
+            tokenizer_name_or_path,
+            cache_dir=model_args.cache_dir,
+            use_fast=True,
+            revision=model_args.model_revision,
+            use_auth_token=True if model_args.use_auth_token else None,
+        )
+    tokenizer.model_max_length = 512
+
+    model = AutoModelForTokenClassification.from_pretrained(
+        model_args.model_name_or_path,
+        from_tf=bool(".ckpt" in 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,
+    )
+
+    # Tokenizer check: this script requires a fast tokenizer.
+    if not isinstance(tokenizer, PreTrainedTokenizerFast):
+        raise ValueError(
+            "This example script only works for models that have a fast tokenizer. Checkout the big table of models "
+            "at https://huggingface.co/transformers/index.html#supported-frameworks to find the model types that meet this "
+            "requirement"
+        )
+
+    # Preprocessing the dataset
+    # Padding strategy
+    padding = "max_length" if data_args.pad_to_max_length else False
+
+    # Tokenize all texts and align the labels with them.
+    def tokenize_and_align_labels(examples):
+        tokenized_inputs = tokenizer(
+            examples[text_column_name],
+            padding=padding,
+            max_length=512,
+            truncation=True,
+            # We use this argument because the texts in our dataset are lists of words (with a label for each word).
+            is_split_into_words=True,
+        )
+        labels = []
+        for i, label in enumerate(examples[label_column_name]):
+            word_ids = tokenized_inputs.word_ids(batch_index=i)
+            previous_word_idx = None
+            label_ids = []
+            for word_idx in word_ids:
+                # Special tokens have a word id that is None. We set the label to -100 so they are automatically
+                # ignored in the loss function.
+                if word_idx is None:
+                    label_ids.append(-100)
+                # We set the label for the first token of each word.
+                elif word_idx != previous_word_idx:
+                    label_ids.append(label_to_id[label[word_idx]])
+                # For the other tokens in a word, we set the label to either the current label or -100, depending on
+                # the label_all_tokens flag.
+                else:
+                    label_ids.append(label_to_id[label[word_idx]] if data_args.label_all_tokens else -100)
+                previous_word_idx = word_idx
+
+            labels.append(label_ids)
+        tokenized_inputs["labels"] = labels
+        return tokenized_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(
+                tokenize_and_align_labels,
+                batched=True,
+                num_proc=data_args.preprocessing_num_workers,
+                load_from_cache_file=not data_args.overwrite_cache,
+                desc="Running tokenizer on train dataset",
+            )
+
+    if training_args.do_eval:
+        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(
+                tokenize_and_align_labels,
+                batched=True,
+                num_proc=data_args.preprocessing_num_workers,
+                load_from_cache_file=not data_args.overwrite_cache,
+                desc="Running tokenizer on validation dataset",
+            )
+
+    if training_args.do_predict:
+        if "test" not in raw_datasets:
+            raise ValueError("--do_predict requires a test dataset")
+        predict_dataset = raw_datasets["test"]
+        if data_args.max_predict_samples is not None:
+            predict_dataset = predict_dataset.select(range(data_args.max_predict_samples))
+        with training_args.main_process_first(desc="prediction dataset map pre-processing"):
+            predict_dataset = predict_dataset.map(
+                tokenize_and_align_labels,
+                batched=True,
+                num_proc=data_args.preprocessing_num_workers,
+                load_from_cache_file=not data_args.overwrite_cache,
+                desc="Running tokenizer on prediction dataset",
+            )
+
+    # Data collator
+    data_collator = DataCollatorForTokenClassification(tokenizer, pad_to_multiple_of=8 if training_args.fp16 else None)
+
+    # Metrics
+    metric = load_metric("seqeval")
+
+    def compute_metrics(p):
+        predictions, labels = p
+        predictions = np.argmax(predictions, axis=2)
+
+        # Remove ignored index (special tokens)
+        true_predictions = [
+            [label_list[p] for (p, l) in zip(prediction, label) if l != -100]
+            for prediction, label in zip(predictions, labels)
+        ]
+        true_labels = [
+            [label_list[l] for (p, l) in zip(prediction, label) if l != -100]
+            for prediction, label in zip(predictions, labels)
+        ]
+
+        results = metric.compute(predictions=true_predictions, references=true_labels)
+        if data_args.return_entity_level_metrics:
+            # Unpack nested dictionaries
+            final_results = {}
+            for key, value in results.items():
+                if isinstance(value, dict):
+                    for n, v in value.items():
+                        final_results[f"{key}_{n}"] = v
+                else:
+                    final_results[key] = value
+            return final_results
+        else:
+            return {
+                "precision": results["overall_precision"],
+                "recall": results["overall_recall"],
+                "f1": results["overall_f1"],
+                "accuracy": results["overall_accuracy"],
+            }
+
+    # Initialize our Trainer
+    training_args.run_name = run_name
+    trainer = Trainer(
+        model=model,
+        args=training_args,
+        train_dataset=train_dataset if training_args.do_train else None,
+        eval_dataset=eval_dataset if training_args.do_eval else None,
+        tokenizer=tokenizer,
+        data_collator=data_collator,
+        compute_metrics=compute_metrics,
+    )
+
+    # Training
+    if training_args.do_train:
+        checkpoint = None
+        if training_args.resume_from_checkpoint is not None:
+            checkpoint = training_args.resume_from_checkpoint
+        elif last_checkpoint is not None:
+            checkpoint = last_checkpoint
+        train_result = trainer.train(resume_from_checkpoint=checkpoint)
+        metrics = train_result.metrics
+        trainer.save_model()  # Saves the tokenizer too for easy upload
+
+        max_train_samples = (
+            data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset)
+        )
+        metrics["train_samples"] = min(max_train_samples, len(train_dataset))
+
+        trainer.log_metrics("train", metrics)
+        trainer.save_metrics("train", metrics)
+        trainer.save_state()
+
+    # Evaluation
+    if training_args.do_eval:
+        logger.info("*** Evaluate ***")
+
+        metrics = trainer.evaluate()
+
+        max_eval_samples = data_args.max_eval_samples if data_args.max_eval_samples is not None else len(eval_dataset)
+        metrics["eval_samples"] = min(max_eval_samples, len(eval_dataset))
+
+        trainer.log_metrics("eval", metrics)
+        trainer.save_metrics("eval", metrics)
+
+    # Predict
+    if training_args.do_predict:
+        logger.info("*** Predict ***")
+
+        predictions, labels, metrics = trainer.predict(predict_dataset, metric_key_prefix="predict")
+        predictions = np.argmax(predictions, axis=2)
+
+        # Remove ignored index (special tokens)
+        true_predictions = [
+            [label_list[p] for (p, l) in zip(prediction, label) if l != -100]
+            for prediction, label in zip(predictions, labels)
+        ]
+
+        trainer.log_metrics("predict", metrics)
+        trainer.save_metrics("predict", metrics)
+
+        # Save predictions
+        output_predictions_file = os.path.join(training_args.output_dir, "predictions.txt")
+        if trainer.is_world_process_zero():
+            with open(output_predictions_file, "w") as writer:
+                for prediction in true_predictions:
+                    writer.write(" ".join(prediction) + "\n")
+
+    if training_args.push_to_hub:
+        kwargs = {"finetuned_from": model_args.model_name_or_path, "tasks": "token-classification"}
+        if data_args.dataset_name is not None:
+            kwargs["dataset_tags"] = data_args.dataset_name
+            if data_args.dataset_config_name is not None:
+                kwargs["dataset_args"] = data_args.dataset_config_name
+                kwargs["dataset"] = f"{data_args.dataset_name} {data_args.dataset_config_name}"
+            else:
+                kwargs["dataset"] = data_args.dataset_name
+
+        trainer.push_to_hub(**kwargs)
+
+
+def _mp_fn(index):
+    # For xla_spawn (TPUs)
+    main()
+
+
+if __name__ == "__main__":
+    main()

evaluation/xnli.yaml

@@ -0,0 +1,55 @@
+name: BERTIN XNLI es
+project: bertin-eval
+enitity: versae
+program: run_glue.py
+command:
+  - ${env}
+  - ${interpreter}
+  - ${program}
+  - ${args}
+method: grid
+metric:
+  name: eval/accuracy
+  goal: maximize
+parameters:
+  model_name_or_path:
+    values:
+    - bertin-project/bertin-base-gaussian-exp-512seqlen
+    - bertin-project/bertin-base-random-exp-512seqlen
+    - bertin-project/bertin-base-gaussian
+    - bertin-project/bertin-base-stepwise
+    - bertin-project/bertin-base-random
+    - bertin-project/bertin-roberta-base-spanish
+    - flax-community/bertin-roberta-large-spanish
+    - BSC-TeMU/roberta-base-bne
+    - dccuchile/bert-base-spanish-wwm-cased
+    - bert-base-multilingual-cased
+  num_train_epochs:
+    values: [5]
+  task_name:
+    value: xnli
+  dataset_name:
+    value: xnli
+  dataset_config_name:
+    value: es
+  output_dir:
+    value: ./outputs
+  overwrite_output_dir:
+    value: true
+  resume_from_checkpoint:
+    value: false
+  max_seq_length:
+    value: 512
+  pad_to_max_length:
+    value: true
+  per_device_train_batch_size:
+    value: 16
+  per_device_eval_batch_size:
+    value: 16
+  save_total_limit:
+    value: 1
+  do_train:
+    value: true
+  do_eval:
+    value: true
+

mc4/mc4.py

@@ -376,13 +376,13 @@ class Mc4(datasets.GeneratorBasedBuilder):
                 for lang in self.config.languages
                 for index in range(_N_SHARDS_PER_SPLIT[lang][split])
             ]
-        if "train" in self.data_files:
+        if self.data_files and "train" in self.data_files:
             train_downloaded_files = self.data_files["train"]
             if not isinstance(train_downloaded_files, (tuple, list)):
                 train_downloaded_files = [train_downloaded_files]
         else:
             train_downloaded_files = dl_manager.download(data_urls["train"])
-        if "validation" in self.data_files:
+        if self.data_files and "validation" in self.data_files:
             validation_downloaded_files = self.data_files["validation"]
             if not isinstance(validation_downloaded_files, (tuple, list)):
                 validation_downloaded_files = [validation_downloaded_files]
@@ -417,7 +417,7 @@ class Mc4(datasets.GeneratorBasedBuilder):
                                 if self.should_keep_doc(
                                     example["text"],
                                     factor=self.sampling_factor,
-                                    boundaries=self.boundaries
+                                    boundaries=self.boundaries,
                                     **self.kwargs):
                                     yield id_, example
                                     id_ += 1