run_finetuning_QA.py

# coding=utf-8
""" Fine-tuning BERT on question answering tasks."""

from __future__ import absolute_import, division, print_function

import argparse
import logging
import os
import random
import glob
import timeit
import json

import numpy as np
import torch
from torch.utils.data import (DataLoader, RandomSampler, SequentialSampler,
                              TensorDataset, Subset)
from torch.utils.data.distributed import DistributedSampler

try:
    from torch.utils.tensorboard import SummaryWriter
except:
    from tensorboardX import SummaryWriter

from tqdm import tqdm, trange

from transformers import (WEIGHTS_NAME, BertConfig, BertForQuestionAnswering, BertTokenizer)

from transformers import AdamW, get_linear_schedule_with_warmup

from ..utils import utils_squad
from utils_squad import (read_squad_examples, convert_examples_to_features,
                         RawResult, write_predictions)

# The follwing import is the official SQuAD evaluation script (2.0).
# You can remove it from the dependencies if you are using this script outside of the library
# We've added it here for automated tests (see examples/test_examples.py file)
from utils_squad_evaluate import EVAL_OPTS, main as evaluate_on_squad

logger = logging.getLogger(__name__)

ALL_MODELS = sum((tuple(conf.pretrained_config_archive_map.keys()) \
                  for conf in (BertConfig, )), ())

MODEL_CLASSES = {
    'bert': (BertConfig, BertForQuestionAnswering, BertTokenizer),
}


def set_seed(args):
    random.seed(args.seed)
    np.random.seed(args.seed)
    torch.manual_seed(args.seed)
    if args.n_gpu > 0:
        torch.cuda.manual_seed_all(args.seed)

        
def to_list(tensor):
    return tensor.detach().cpu().tolist()

  
def train(args, train_dataset, model, tokenizer):
    """ Train the model """
    if args.local_rank in [-1, 0]:
        tb_writer = SummaryWriter()

    args.train_batch_size = args.per_gpu_train_batch_size * max(1, args.n_gpu)
    train_sampler = RandomSampler(train_dataset) if args.local_rank == -1 else DistributedSampler(train_dataset)
    train_dataloader = DataLoader(train_dataset, sampler=train_sampler, batch_size=args.train_batch_size)

    if args.max_steps > 0:
        t_total = args.max_steps
        args.num_train_epochs = args.max_steps // (len(train_dataloader) // args.gradient_accumulation_steps) + 1
    else:
        t_total = len(train_dataloader) // args.gradient_accumulation_steps * args.num_train_epochs

    # Prepare optimizer and schedule (linear warmup and decay)
    no_decay = ['bias', 'LayerNorm.weight']

    if args.finetune_feature_extractor:
        optimizer_grouped_parameters = [
            {'params': [p for n, p in model.named_parameters() if not any(nd in n for nd in no_decay)],
             'weight_decay': args.weight_decay},
            {'params': [p for n, p in model.named_parameters() if any(nd in n for nd in no_decay)], 'weight_decay': 0.0}
        ]
    else:
        optimizer_grouped_parameters = [
            {'params': [p for n, p in model.named_parameters() if not any(nd in n for nd in no_decay)
                        and args.model_type not in n],
             'weight_decay': args.weight_decay},
            {'params': [p for n, p in model.named_parameters() if any(nd in n for nd in no_decay)
                        and args.model_type not in n],
             'weight_decay': 0.0}
        ]

    optimizer = AdamW(optimizer_grouped_parameters, lr=args.learning_rate, eps=args.adam_epsilon)
    scheduler = get_linear_schedule_with_warmup(optimizer, num_warmup_steps=args.warmup_steps, num_training_steps=t_total)
    if args.fp16:
        try:
            from apex import amp
        except ImportError:
            raise ImportError("Please install apex from https://www.github.com/nvidia/apex to use fp16 training.")
        model, optimizer = amp.initialize(model, optimizer, opt_level=args.fp16_opt_level)

    # multi-gpu training (should be after apex fp16 initialization)
    if args.n_gpu > 1:
        model = torch.nn.DataParallel(model)

    # Distributed training (should be after apex fp16 initialization)
    if args.local_rank != -1:
        model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.local_rank],
                                                          output_device=args.local_rank,
                                                          find_unused_parameters=True)

    # Train!
    logger.info("***** Running training *****")
    logger.info("  Num examples = %d", len(train_dataset))
    logger.info("  Num Epochs = %d", args.num_train_epochs)
    logger.info("  Instantaneous batch size per GPU = %d", args.per_gpu_train_batch_size)
    logger.info("  Total train batch size (w. parallel, distributed & accumulation) = %d",
                   args.train_batch_size * args.gradient_accumulation_steps * (torch.distributed.get_world_size() if args.local_rank != -1 else 1))
    logger.info("  Gradient Accumulation steps = %d", args.gradient_accumulation_steps)
    logger.info("  Total optimization steps = %d", t_total)

    global_step = 0
    tr_loss, logging_loss = 0.0, 0.0
    model.zero_grad()
    train_iterator = trange(int(args.num_train_epochs), desc="Epoch", disable=args.local_rank not in [-1, 0])
    set_seed(args)  # Added here for reproductibility (even between python 2 and 3)
    epoch = 0

    if args.local_rank in [-1, 0]:
        # Save model checkpoint
        if args.save == 'all' or args.save == str(epoch):
            checkpoint_output_dir = os.path.join(args.output_dir, 'checkpoint-{}'.format(epoch))
            if not os.path.exists(checkpoint_output_dir):
                os.makedirs(checkpoint_output_dir)
            model_to_save = model.module if hasattr(model,
                                                    'module') else model  # Take care of distributed/parallel training
            model_to_save.save_pretrained(checkpoint_output_dir)
            torch.save(args, os.path.join(checkpoint_output_dir, 'training_args.bin'))
            tokenizer.save_pretrained(checkpoint_output_dir)
            logger.info("Saving model checkpoint to %s", checkpoint_output_dir)

    for _ in train_iterator:
        epoch += 1
        epoch_iterator = tqdm(train_dataloader, desc="Iteration", disable=args.local_rank not in [-1, 0])
        for step, batch in enumerate(epoch_iterator):
            model.train()
            batch = tuple(t.to(args.device) for t in batch)
            inputs = {'input_ids':       batch[0],
                      'attention_mask':  batch[1],
                      'token_type_ids':  batch[2],
                      'start_positions': batch[3],
                      'end_positions':   batch[4]}
            outputs = model(**inputs)
            loss = outputs[0]  # model outputs are always tuple in transformers (see doc)

            if args.n_gpu > 1:
                loss = loss.mean() # mean() to average on multi-gpu parallel (not distributed) training
            if args.gradient_accumulation_steps > 1:
                loss = loss / args.gradient_accumulation_steps

            if args.fp16:
                with amp.scale_loss(loss, optimizer) as scaled_loss:
                    scaled_loss.backward()
            else:
                loss.backward()

            tr_loss += loss.item()
            if (step + 1) % args.gradient_accumulation_steps == 0:
                if args.fp16:
                    torch.nn.utils.clip_grad_norm_(amp.master_params(optimizer), args.max_grad_norm)
                else:
                    torch.nn.utils.clip_grad_norm_(model.parameters(), args.max_grad_norm)

                optimizer.step()
                scheduler.step()  # Update learning rate schedule
                model.zero_grad()
                global_step += 1

                if args.local_rank in [-1, 0] and args.logging_steps > 0 and global_step % args.logging_steps == 0:
                    # Log metrics
                    if args.local_rank == -1 and args.evaluate_during_training:  # Only evaluate when single GPU otherwise metrics may not average well
                        results = evaluate(args, model, tokenizer)
                        for key, value in results.items():
                            tb_writer.add_scalar('eval_{}'.format(key), value, global_step)
                    tb_writer.add_scalar('lr', scheduler.get_lr()[0], global_step)
                    tb_writer.add_scalar('loss', (tr_loss - logging_loss)/args.logging_steps, global_step)
                    logging_loss = tr_loss

            if args.max_steps > 0 and global_step > args.max_steps:
                epoch_iterator.close()
                break

        if args.local_rank in [-1, 0]:
            # Save model checkpoint
            if args.save == 'all' or args.save == str(epoch):
                checkpoint_output_dir = os.path.join(args.output_dir, 'checkpoint-{}'.format(epoch))
                if not os.path.exists(checkpoint_output_dir):
                    os.makedirs(checkpoint_output_dir)
                model_to_save = model.module if hasattr(model,
                                                        'module') else model  # Take care of distributed/parallel training
                model_to_save.save_pretrained(checkpoint_output_dir)
                torch.save(args, os.path.join(checkpoint_output_dir, 'training_args.bin'))
                tokenizer.save_pretrained(checkpoint_output_dir)
                logger.info("Saving model checkpoint to %s", checkpoint_output_dir)

        if args.max_steps > 0 and global_step > args.max_steps:
            train_iterator.close()
            break

    if args.local_rank in [-1, 0]:
        tb_writer.close()

    return global_step, tr_loss / global_step


def evaluate(args, model, tokenizer, checkpoint_id=None, prefix=""):
    dataset, examples, features = load_and_cache_examples(args, tokenizer, evaluate=True, output_examples=True)
    if args.eval_data_subset > 0:
        dataset = Subset(dataset, list(range(min(args.eval_data_subset, len(dataset)))))

    if not os.path.exists(args.output_dir) and args.local_rank in [-1, 0]:
        os.makedirs(args.output_dir)

    args.eval_batch_size = args.per_gpu_eval_batch_size * max(1, args.n_gpu)
    # Note that DistributedSampler samples randomly
    eval_sampler = SequentialSampler(dataset) if args.local_rank == -1 else DistributedSampler(dataset)
    eval_dataloader = DataLoader(dataset, sampler=eval_sampler, batch_size=args.eval_batch_size)

    # Eval!
    logger.info("***** Running evaluation {} *****".format(prefix))
    logger.info("  Num examples = %d", len(dataset))
    logger.info("  Batch size = %d", args.eval_batch_size)
    all_results = []
    start_time = timeit.default_timer()
    for batch in tqdm(eval_dataloader, desc="Evaluating"):
        model.eval()
        batch = tuple(t.to(args.device) for t in batch)
        with torch.no_grad():
            inputs = {'input_ids':      batch[0],
                      'attention_mask': batch[1],
                      'token_type_ids':  batch[2],
                      }
            example_indices = batch[3]
            outputs = model(**inputs)

        for i, example_index in enumerate(example_indices):
            eval_feature = features[example_index.item()]
            unique_id = int(eval_feature.unique_id)
            result = RawResult(unique_id    = unique_id,
                               start_logits = to_list(outputs[0][i]),
                               end_logits   = to_list(outputs[1][i]))
            all_results.append(result)

    evalTime = timeit.default_timer() - start_time
    logger.info("  Evaluation done in total %f secs (%f sec per example)", evalTime, evalTime / len(dataset))

    # Compute predictions
    output_prediction_file = os.path.join(args.output_dir,
                                          "predictions_{}.json".format(checkpoint_id if checkpoint_id is not None else prefix))
    output_nbest_file = os.path.join(args.output_dir,
                                     "nbest_predictions_{}.json".format(checkpoint_id if checkpoint_id is not None else prefix))
    if args.version_2_with_negative:
        output_null_log_odds_file = os.path.join(args.output_dir,
                                                 "null_odds_{}.json".format(checkpoint_id if checkpoint_id is not None else prefix))
    else:
        output_null_log_odds_file = None

    write_predictions(examples, features, all_results, args.n_best_size,
                    args.max_answer_length, args.do_lower_case, output_prediction_file,
                    output_nbest_file, output_null_log_odds_file, args.verbose_logging,
                    args.version_2_with_negative, args.null_score_diff_threshold)

    # Evaluate with the official SQuAD script
    evaluate_options = EVAL_OPTS(data_file=args.predict_file,
                                 pred_file=output_prediction_file,
                                 na_prob_file=output_null_log_odds_file)
    results = evaluate_on_squad(evaluate_options)

    output_eval_file = os.path.join(args.output_dir,
                                    "eval_results_{}.txt".format(checkpoint_id) if checkpoint_id is not None
                                    else "eval_results.txt"
                                    )
    with open(output_eval_file, "w") as writer:
        logger.info("***** Eval results {} *****".format(prefix))
        for key in sorted(results.keys()):
            logger.info("  %s = %s", key, str(results[key]))
            writer.write("%s = %s\n" % (key, str(results[key])))

    return results


def load_and_cache_examples(args, tokenizer, evaluate=False, output_examples=False):
    if args.local_rank not in [-1, 0] and not evaluate:
        torch.distributed.barrier()  # Make sure only the first process in distributed training process the dataset, and the others will use the cache

    # Load data features from cache or dataset file
    input_file = args.predict_file if evaluate else args.train_file
    cached_features_file = os.path.join(os.path.dirname(input_file), 'cached_{}_{}_{}'.format(
        'dev' if evaluate else 'train',
        list(filter(None, args.model_name_or_path.split('/'))).pop(),
        str(args.max_seq_length)))
    if os.path.exists(cached_features_file) and not args.overwrite_cache and not output_examples:
        logger.info("Loading features from cached file %s", cached_features_file)
        features = torch.load(cached_features_file)
    else:
        logger.info("Creating features from dataset file at %s", input_file)
        examples = read_squad_examples(input_file=input_file, 
                                       is_training=not evaluate,
                                       version_2_with_negative=args.version_2_with_negative)
        features = convert_examples_to_features(examples=examples,
                                                tokenizer=tokenizer,
                                                max_seq_length=args.max_seq_length,
                                                doc_stride=args.doc_stride,
                                                max_query_length=args.max_query_length,
                                                is_training=not evaluate,
                                                cls_token_segment_id=0,
                                                pad_token_segment_id=0,
                                                cls_token_at_end=False,
                                                sequence_a_is_doc=False)
        if args.local_rank in [-1, 0]:
            logger.info("Saving features into cached file %s", cached_features_file)
            torch.save(features, cached_features_file)

    if args.local_rank == 0 and not evaluate:
        torch.distributed.barrier()  # Make sure only the first process in distributed training process the dataset, and the others will use the cache

    # Convert to Tensors and build dataset
    all_input_ids = torch.tensor([f.input_ids for f in features], dtype=torch.long)
    all_input_mask = torch.tensor([f.input_mask for f in features], dtype=torch.long)
    all_segment_ids = torch.tensor([f.segment_ids for f in features], dtype=torch.long)
    all_cls_index = torch.tensor([f.cls_index for f in features], dtype=torch.long)
    all_p_mask = torch.tensor([f.p_mask for f in features], dtype=torch.float)
    if evaluate:
        all_example_index = torch.arange(all_input_ids.size(0), dtype=torch.long)
        dataset = TensorDataset(all_input_ids, all_input_mask, all_segment_ids,
                                all_example_index, all_cls_index, all_p_mask)
    else:
        all_start_positions = torch.tensor([f.start_position for f in features], dtype=torch.long)
        all_end_positions = torch.tensor([f.end_position for f in features], dtype=torch.long)
        dataset = TensorDataset(all_input_ids, all_input_mask, all_segment_ids,
                                all_start_positions, all_end_positions,
                                all_cls_index, all_p_mask)

    if output_examples:
        return dataset, examples, features
    return dataset


def main():
    parser = argparse.ArgumentParser()

    ## Required parameters
    parser.add_argument("--train_file", default=None, type=str, required=True,
                        help="SQuAD json for training. E.g., train-v1.1.json")
    parser.add_argument("--predict_file", default=None, type=str, required=True,
                        help="SQuAD json for predictions. E.g., dev-v1.1.json or test-v1.1.json")
    parser.add_argument("--model_type", default=None, type=str, required=True,
                        help="Model type selected in the list: " + ", ".join(MODEL_CLASSES.keys()))
    parser.add_argument("--model_name_or_path", default=None, type=str, required=True,
                        help="Path to pre-trained model or shortcut name selected in the list: " + ", ".join(ALL_MODELS))
    parser.add_argument("--output_dir", default=None, type=str, required=True,
                        help="The output directory where the model checkpoints and predictions will be written.")

    ## Other parameters
    parser.add_argument("--config_name", default="", type=str,
                        help="Pretrained config name or path if not the same as model_name")
    parser.add_argument("--tokenizer_name", default="", type=str,
                        help="Pretrained tokenizer name or path if not the same as model_name")
    parser.add_argument("--cache_dir", default="", type=str,
                        help="Where do you want to store the pre-trained models downloaded from s3")

    parser.add_argument('--version_2_with_negative', default=False, type=eval,
                        help='If true, the SQuAD examples contain some that do not have an answer.')
    parser.add_argument('--null_score_diff_threshold', type=float, default=0.0,
                        help="If null_score - best_non_null is greater than the threshold predict null.")
    parser.add_argument("--train_data_subset", type=int, default=-1,
                        help="If > 0: limit the training data to a subset of train_data_subset instances.")
    parser.add_argument("--eval_data_subset", type=int, default=-1,
                        help="If > 0: limit the evaluation data to a subset of eval_data_subset instances.")
    parser.add_argument("--max_seq_length", default=384, type=int,
                        help="The maximum total input sequence length after WordPiece tokenization. Sequences "
                             "longer than this will be truncated, and sequences shorter than this will be padded.")
    parser.add_argument("--doc_stride", default=128, type=int,
                        help="When splitting up a long document into chunks, how much stride to take between chunks.")
    parser.add_argument("--max_query_length", default=64, type=int,
                        help="The maximum number of tokens for the question. Questions longer than this will "
                             "be truncated to this length.")
    parser.add_argument("--do_train", action='store_true',
                        help="Whether to run training.")
    parser.add_argument("--do_eval", action='store_true',
                        help="Whether to run eval on the dev set.")
    parser.add_argument("--evaluate_during_training", action='store_true',
                        help="Rul evaluation during training at each logging step.")
    parser.add_argument("--do_lower_case", action='store_true',
                        help="Set this flag if you are using an uncased model.")
    parser.add_argument("--finetune_feature_extractor", default=True, type=eval,
                        help="Whether to fine-tune the feature extractor.")

    parser.add_argument("--per_gpu_train_batch_size", default=8, type=int,
                        help="Batch size per GPU/CPU for training.")
    parser.add_argument("--per_gpu_eval_batch_size", default=8, type=int,
                        help="Batch size per GPU/CPU for evaluation.")
    parser.add_argument("--learning_rate", default=5e-5, type=float,
                        help="The initial learning rate for Adam.")
    parser.add_argument('--gradient_accumulation_steps', type=int, default=1,
                        help="Number of updates steps to accumulate before performing a backward/update pass.")
    parser.add_argument("--weight_decay", default=0.0, type=float,
                        help="Weight deay if we apply some.")
    parser.add_argument("--adam_epsilon", default=1e-8, type=float,
                        help="Epsilon for Adam optimizer.")
    parser.add_argument("--max_grad_norm", default=1.0, type=float,
                        help="Max gradient norm.")
    parser.add_argument("--num_train_epochs", default=3.0, type=float,
                        help="Total number of training epochs to perform.")
    parser.add_argument("--max_steps", default=-1, type=int,
                        help="If > 0: set total number of training steps to perform. Override num_train_epochs.")
    parser.add_argument("--warmup_steps", default=0, type=int,
                        help="Linear warmup over warmup_steps.")
    parser.add_argument("--n_best_size", default=20, type=int,
                        help="The total number of n-best predictions to generate in the nbest_predictions.json output file.")
    parser.add_argument("--max_answer_length", default=30, type=int,
                        help="The maximum length of an answer that can be generated. This is needed because the start "
                             "and end predictions are not conditioned on one another.")
    parser.add_argument("--verbose_logging", action='store_true',
                        help="If true, all of the warnings related to data processing will be printed. "
                             "A number of warnings are expected for a normal SQuAD evaluation.")

    parser.add_argument('--logging_steps', type=int, default=50,
                        help="Log every X updates steps.")
    parser.add_argument("--eval_all_checkpoints", action='store_true',
                        help="Evaluate all checkpoints starting with the same prefix as model_name ending and ending with step number")
    parser.add_argument("--no_cuda", action='store_true',
                        help="Whether not to use CUDA when available")
    parser.add_argument('--overwrite_output_dir', action='store_true',
                        help="Overwrite the content of the output directory")
    parser.add_argument('--overwrite_cache', action='store_true',
                        help="Overwrite the cached training and evaluation sets")
    parser.add_argument('--seed', type=int, default=42,
                        help="random seed for initialization")
    parser.add_argument('--model_load_mode', type=str, default='all',
                        help="Select load mode from ['base_model_only', 'all']")
    parser.add_argument('--save', type=str, default='all',
                        help="Select load mode from ['all', '0', '1', '2', '3', ...]")

    parser.add_argument("--local_rank", type=int, default=-1,
                        help="local_rank for distributed training on gpus")
    parser.add_argument('--fp16', action='store_true',
                        help="Whether to use 16-bit (mixed) precision (through NVIDIA apex) instead of 32-bit")
    parser.add_argument('--fp16_opt_level', type=str, default='O1',
                        help="For fp16: Apex AMP optimization level selected in ['O0', 'O1', 'O2', and 'O3']."
                             "See details at https://nvidia.github.io/apex/amp.html")
    parser.add_argument('--server_ip', type=str, default='', help="Can be used for distant debugging.")
    parser.add_argument('--server_port', type=str, default='', help="Can be used for distant debugging.")
    args = parser.parse_args()

    if os.path.exists(args.output_dir) and os.listdir(args.output_dir) and args.do_train and not args.overwrite_output_dir:
        raise ValueError("Output directory ({}) already exists and is not empty. Use --overwrite_output_dir to overcome.".format(args.output_dir))

    # Create output directory if needed
    if not os.path.exists(args.output_dir):
        os.makedirs(args.output_dir)
    with open(os.path.join(args.output_dir, 'run_args.txt'), 'w') as f:
        f.write(json.dumps(args.__dict__, indent=2))
        f.close()

    # Setup distant debugging if needed
    if args.server_ip and args.server_port:
        # Distant debugging - see https://code.visualstudio.com/docs/python/debugging#_attach-to-a-local-script
        import ptvsd
        print("Waiting for debugger attach")
        ptvsd.enable_attach(address=(args.server_ip, args.server_port), redirect_output=True)
        ptvsd.wait_for_attach()

    # Setup CUDA, GPU & distributed training
    if args.local_rank == -1 or args.no_cuda:
        device = torch.device("cuda" if torch.cuda.is_available() and not args.no_cuda else "cpu")
        args.n_gpu = torch.cuda.device_count()
    else:  # Initializes the distributed backend which will take care of sychronizing nodes/GPUs
        torch.cuda.set_device(args.local_rank)
        device = torch.device("cuda", args.local_rank)
        torch.distributed.init_process_group(backend='nccl')
        args.n_gpu = 1
    args.device = device

    # Setup logging
    logging.basicConfig(format = '%(asctime)s - %(levelname)s - %(name)s -   %(message)s',
                        datefmt = '%m/%d/%Y %H:%M:%S',
                        level = logging.INFO if args.local_rank in [-1, 0] else logging.WARN)
    logger.warning("Process rank: %s, device: %s, n_gpu: %s, distributed training: %s, 16-bits training: %s",
                    args.local_rank, device, args.n_gpu, bool(args.local_rank != -1), args.fp16)

    # Set seed
    set_seed(args)
    num_labels = 2
    # Load pretrained model and tokenizer
    if args.local_rank not in [-1, 0]:
        torch.distributed.barrier()  # Make sure only the first process in distributed training will download model & vocab

    if args.model_load_mode not in ['base_model_only', 'all']:
        raise ValueError("Model load mode not found: %s" % (args.model_load_mode))

    state_dict_with_prefix = None
    if args.model_load_mode == 'base_model_only':
        archive_file = os.path.join(args.model_name_or_path, WEIGHTS_NAME)
        model_state_dict = torch.load(archive_file)
        state_dict_with_prefix = {}
        for key, value in model_state_dict.items():
            if key.startswith(args.model_type):
                state_dict_with_prefix[key] = value

    args.model_type = args.model_type.lower()
    config_class, model_class, tokenizer_class = MODEL_CLASSES[args.model_type]
    config = config_class.from_pretrained(args.config_name if args.config_name else args.model_name_or_path,
                                          num_labels=num_labels,
                                          cache_dir=args.cache_dir if args.cache_dir else None)
    tokenizer = tokenizer_class.from_pretrained(args.tokenizer_name if args.tokenizer_name else args.model_name_or_path,
                                                do_lower_case=args.do_lower_case,
                                                cache_dir=args.cache_dir if args.cache_dir else None)
    model = model_class.from_pretrained(args.model_name_or_path,
                                        from_tf=bool('.ckpt' in args.model_name_or_path),
                                        config=config,
                                        cache_dir=args.cache_dir if args.cache_dir else None,
                                        state_dict=state_dict_with_prefix)

    if args.local_rank == 0:
        torch.distributed.barrier()  # Make sure only the first process in distributed training will download model & vocab

    model.to(args.device)

    logger.info("Training/evaluation parameters %s", args)
    # Good practice: save your training arguments together with the trained model
    torch.save(args, os.path.join(args.output_dir, 'training_args.bin'))
    tokenizer.save_pretrained(args.output_dir)

    # Before we do anything with models, we want to ensure that we get fp16 execution of torch.einsum if args.fp16 is set.
    # Otherwise it'll default to "promote" mode, and we'll get fp32 operations. Note that running `--fp16_opt_level="O2"` will
    # remove the need for this code, but it is still valid.
    if args.fp16:
        try:
            import apex
            apex.amp.register_half_function(torch, 'einsum')
        except ImportError:
            raise ImportError("Please install apex from https://www.github.com/nvidia/apex to use fp16 training.")

    # Training
    if args.do_train:
        train_dataset = load_and_cache_examples(args, tokenizer, evaluate=False, output_examples=False)
        if args.train_data_subset > 0:
            train_dataset = Subset(train_dataset, list(range(min(args.train_data_subset, len(train_dataset)))))
        global_step, tr_loss = train(args, train_dataset, model, tokenizer)
        logger.info(" global_step = %s, average loss = %s", global_step, tr_loss)

    # Evaluation - we can ask to evaluate all the checkpoints (sub-directories) in a directory
    results = {}
    if args.do_eval and args.local_rank in [-1, 0]:
        eval_list = []
        if args.save == 'all':
            eval_list = range(int(args.num_train_epochs + 1))
        elif str(args.save).isdigit():
            eval_list = [int(args.save)]
        if len(eval_list) > 0:
            for epoch in eval_list:
                checkpoints = [os.path.join(args.output_dir, 'checkpoint-{}'.format(epoch))]
                if args.eval_all_checkpoints:
                    checkpoints = list(os.path.dirname(c) for c in sorted(glob.glob(args.output_dir + '/**/' + WEIGHTS_NAME, recursive=True)))
                    logging.getLogger("transformers.modeling_utils").setLevel(logging.WARN)  # Reduce model loading logs
                logger.info("Evaluate the following checkpoints: %s", checkpoints)
                for checkpoint in checkpoints:
                    global_step = checkpoint.split('-')[-1] if len(checkpoints) > 1 else ""
                    prefix = checkpoint.split('/')[-1] if checkpoint.find('checkpoint') != -1 else ""
                    model = model_class.from_pretrained(checkpoint)
                    tokenizer = tokenizer_class.from_pretrained(args.output_dir, do_lower_case=args.do_lower_case)

                    model.to(args.device)
                    result = evaluate(args, model, tokenizer, checkpoint_id=epoch, prefix=prefix)
                    result = dict((k + '_{}'.format(global_step), v) for k, v in result.items())
                    results.update(result)

    logger.info("Results: {}".format(results))

    return results


if __name__ == "__main__":
    main()