train_soft_classifier.py

# coding=utf-8
# Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team.
# Copyright (c) 2018, NVIDIA CORPORATION.  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 multi-lingual models on XNLI (e.g. Bert, DistilBERT, XLM).
    Adapted from `examples/text-classification/run_glue.py`"""

import argparse
import glob
import json
import logging
import os
import random
import pickle as pk

import numpy as np
import torch
import torch.nn as nn
from sklearn.metrics import f1_score
from torch.nn import CrossEntropyLoss, MSELoss
from torch.utils.data import (DataLoader, RandomSampler, SequentialSampler,
                              TensorDataset)
from torch.utils.data.distributed import DistributedSampler
from tqdm import tqdm, trange
from transformers import (WEIGHTS_NAME, AdamW, AutoConfig,
                          AutoModelForSequenceClassification, AutoTokenizer,
                          BertForSequenceClassification, BertModel,
                          BertPreTrainedModel, DataProcessor, InputExample,
                          get_linear_schedule_with_warmup)
from transformers import \
    glue_convert_examples_to_features as convert_examples_to_features

from preprocessing_utils import load, load_classnames, load_labels, load_probs, load_text
from utils import DATA_FOLDER_PATH, INTERMEDIATE_DATA_FOLDER_PATH


class DProcessor(DataProcessor):
    def __init__(self, task):
        self.task_name = task
        self.train_dir = f"{task}"
        self.test_dir = f"{task}"
        self.labels = list(range(len(load_classnames(os.path.join(DATA_FOLDER_PATH, task)))))

    def get_train_examples(self, data_dir):
        """See base class."""
        examples = []
        labels = []
        for i, (text, label) in enumerate(zip(load_text(os.path.join(data_dir, self.train_dir)),
                                              load_probs(os.path.join(data_dir, self.train_dir)))):
            guid = "%s-%s" % ("train", i)
            # examples.append(InputExample(guid=guid, text_a=text, label=label))
            examples.append(InputExample(guid=guid, text_a=text))
            labels.append(np.array(label))
        return examples, np.array(labels)

    def get_test_examples(self, data_dir, dataset_name):
        """See base class."""
        examples = []
        labels = []
        with open(os.path.join(INTERMEDIATE_DATA_FOLDER_PATH, self.task_name, f"dataset.pk"), "rb") as f:
            dataset = pk.load(f)
        for i, (text, label) in enumerate(zip(dataset["cleaned_text"],
                                              load_labels(os.path.join(data_dir, self.test_dir)))):
            guid = "%s-%s" % ("test", i)
            # examples.append(InputExample(guid=guid, text_a=text, label=label))
            examples.append(InputExample(guid=guid, text_a=text))
            temp = np.zeros((len(self.labels)))
            temp[int(label)] = 1
            labels.append(np.copy(temp))
        return examples, np.array(labels)

    def get_labels(self):
        """See base class."""
        return self.labels


def compute_metrics(task_name, preds, labels):
    return {
        "f1_micro": f1_score(labels, preds, average='micro'),
        "f1_macro": f1_score(labels, preds, average='macro')
    }


class HACK_FOR_DIFFERENT_DATASETS:
    def __init__(self, key):
        self.key = key

    def __getitem__(self, _):
        return self.key

    def __contains__(self, item):
        return True


output_modes = HACK_FOR_DIFFERENT_DATASETS("classification")
processors = HACK_FOR_DIFFERENT_DATASETS(DProcessor)

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

logger = logging.getLogger(__name__)


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 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"]
    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},
    ]
    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
    )
    # NEW: Use KL-Divergence as loss function instead of negative log-likelihood
    kl_loss = nn.KLDivLoss(reduction="batchmean")

    # Check if saved optimizer or scheduler states exist
    if os.path.isfile(os.path.join(args.model_name_or_path, "optimizer.pt")) and os.path.isfile(
            os.path.join(args.model_name_or_path, "scheduler.pt")
    ):
        # Load in optimizer and scheduler states
        optimizer.load_state_dict(torch.load(os.path.join(args.model_name_or_path, "optimizer.pt")))
        scheduler.load_state_dict(torch.load(os.path.join(args.model_name_or_path, "scheduler.pt")))

    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
    epochs_trained = 0
    steps_trained_in_current_epoch = 0
    # Check if continuing training from a checkpoint
    # if os.path.exists(args.model_name_or_path):
    #     # set global_step to gobal_step of last saved checkpoint from model path
    #     global_step = int(args.model_name_or_path.split("-")[-1].split("/")[0])
    #     epochs_trained = global_step // (len(train_dataloader) // args.gradient_accumulation_steps)
    #     steps_trained_in_current_epoch = global_step % (len(train_dataloader) // args.gradient_accumulation_steps)
    #
    #     logger.info("  Continuing training from checkpoint, will skip to saved global_step")
    #     logger.info("  Continuing training from epoch %d", epochs_trained)
    #     logger.info("  Continuing training from global step %d", global_step)
    #     logger.info("  Will skip the first %d steps in the first epoch", steps_trained_in_current_epoch)

    tr_loss, logging_loss = 0.0, 0.0
    model.zero_grad()
    train_iterator = trange(
        epochs_trained, int(args.num_train_epochs), desc="Epoch", disable=args.local_rank not in [-1, 0]
    )
    set_seed(args)  # Added here for reproductibility
    for _ in train_iterator:
        epoch_iterator = tqdm(train_dataloader, desc="Iteration", disable=args.local_rank not in [-1, 0])
        for step, batch in enumerate(epoch_iterator):
            # Skip past any already trained steps if resuming training
            if steps_trained_in_current_epoch > 0:
                steps_trained_in_current_epoch -= 1
                continue

            model.train()
            batch = tuple(t.to(args.device) for t in batch)
            # inputs = {"input_ids": batch[0], "attention_mask": batch[1], "labels": batch[3]}
            inputs = {"input_ids": batch[0], "attention_mask": batch[1]}
            target = batch[3]

            if args.model_type != "distilbert":
                inputs["token_type_ids"] = (
                    batch[2] if args.model_type in ["bert"] else None
                )  # XLM and DistilBERT don't use segment_ids
            logits = model(**inputs)[0]
            # loss = outputs[0]  # model outputs are always tuple in transformers (see doc)
            pred_probs = nn.functional.log_softmax(logits, dim=1)
            loss = kl_loss(pred_probs, target)

            if args.n_gpu > 1:
                loss = loss.mean()  # mean() to average on multi-gpu parallel 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.local_rank in [-1, 0] and args.save_steps > 0 and global_step % args.save_steps == 0:
                    # Save model checkpoint
                    output_dir = os.path.join(args.output_dir, "checkpoint-{}".format(global_step))
                    if not os.path.exists(output_dir):
                        os.makedirs(output_dir)
                    model_to_save = (
                        model.module if hasattr(model, "module") else model
                    )  # Take care of distributed/parallel training
                    model_to_save.save_pretrained(output_dir)
                    tokenizer.save_pretrained(output_dir)

                    torch.save(args, os.path.join(output_dir, "training_args.bin"))
                    logger.info("Saving model checkpoint to %s", output_dir)

                    torch.save(optimizer.state_dict(), os.path.join(output_dir, "optimizer.pt"))
                    torch.save(scheduler.state_dict(), os.path.join(output_dir, "scheduler.pt"))
                    logger.info("Saving optimizer and scheduler states to %s", output_dir)

            if args.max_steps > 0 and global_step > args.max_steps:
                epoch_iterator.close()
                break
        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, prefix="", save_res=False):
    eval_task_names = (args.dataset_name,)
    eval_outputs_dirs = (args.output_dir,)

    results = {}
    for eval_task, eval_output_dir in zip(eval_task_names, eval_outputs_dirs):
        eval_dataset = load_and_cache_examples(args, eval_task, tokenizer, evaluate=True)

        if not os.path.exists(eval_output_dir) and args.local_rank in [-1, 0]:
            os.makedirs(eval_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(eval_dataset)
        eval_dataloader = DataLoader(eval_dataset, sampler=eval_sampler, batch_size=args.eval_batch_size)
        kl_loss = nn.KLDivLoss(reduction="batchmean")

        # multi-gpu eval
        if args.n_gpu > 1 and not isinstance(model, torch.nn.DataParallel):
            model = torch.nn.DataParallel(model)

        # Eval!
        logger.info("***** Running evaluation {} *****".format(prefix))
        logger.info("  Num examples = %d", len(eval_dataset))
        logger.info("  Batch size = %d", args.eval_batch_size)
        eval_loss = 0.0
        nb_eval_steps = 0
        preds = None
        probs = None
        out_label_ids = None
        for batch in tqdm(eval_dataloader, desc="Evaluating"):
            model.eval()
            batch = tuple(t.to(args.device) for t in batch)

            with torch.no_grad():
                # NEW: Converted to soft labels
                # inputs = {"input_ids": batch[0], "attention_mask": batch[1], "labels": batch[3]}
                inputs = {"input_ids": batch[0], "attention_mask": batch[1]}
                target = batch[3]

                if args.model_type != "distilbert":
                    inputs["token_type_ids"] = (
                        batch[2] if args.model_type in ["bert"] else None
                    )  # XLM and DistilBERT don't use segment_ids
                outputs = model(**inputs)
                # tmp_eval_loss, logits = outputs[:2]
                logits = outputs[0]
                pred_probs = nn.functional.log_softmax(logits, dim=1)
                tmp_eval_loss = kl_loss(pred_probs, target)

                eval_loss += tmp_eval_loss.mean().item()
            nb_eval_steps += 1
            if preds is None:
                preds = logits.detach().cpu().numpy()
                probs = nn.functional.softmax(logits, dim=1).detach().cpu().numpy()
                hard_labels = torch.argmax(batch[3], dim=1)
                out_label_ids = hard_labels.detach().cpu().numpy()
            else:
                preds = np.append(preds, logits.detach().cpu().numpy(), axis=0)
                probs = np.append(probs, nn.functional.softmax(logits, dim=1).detach().cpu().numpy(), axis=0)
                hard_labels = torch.argmax(batch[3], dim=1)
                out_label_ids = np.append(out_label_ids, hard_labels.detach().cpu().numpy(), axis=0)

        eval_loss = eval_loss / nb_eval_steps

        # NOTE: (NEW) output the class probabilities into file
        if save_res:
            if os.path.exists(os.path.join(INTERMEDIATE_DATA_FOLDER_PATH, args.dataset_name, f"classifier_results.pk")):
                with open(os.path.join(INTERMEDIATE_DATA_FOLDER_PATH, args.dataset_name, f"classifier_results.pk"), "rb") as f:
                    class_res = pk.load(f)
                class_res["res" + "_" + str(args.iter)] = probs.tolist()
            else:
                class_res = {
                    "res" + "_" + str(args.iter): probs.tolist()
                }
            with open(os.path.join(INTERMEDIATE_DATA_FOLDER_PATH, args.dataset_name, f"classifier_results.pk"), "wb") as f:
                pk.dump(class_res, f)

        # class_probs = os.path.join(INTERMEDIATE_DATA_FOLDER_PATH, args.dataset_name, str(args.confidence_threshold) + "_" + str(args.granularity) + "_probabilities.json")
        # with open(class_probs, "w") as writer:
        #     json.dump(probs.tolist(), writer, indent=2)

        if args.output_mode == "classification":
            preds = np.argmax(preds, axis=1)
        else:
            raise ValueError("No other `output_mode` for XNLI.")
        result = compute_metrics(eval_task, preds, out_label_ids)
        results.update(result)

        output_eval_file = os.path.join(eval_output_dir, prefix, "eval_results.txt")
        with open(output_eval_file, "w") as writer:
            logger.info("***** Eval results {} *****".format(prefix))
            print("***** Eval results *****")
            for key in sorted(result.keys()):
                logger.info("  %s = %s", key, str(result[key]))
                print(key, "=", str(result[key]))
                writer.write("%s = %s\n" % (key, str(result[key])))
        output_labels_file = os.path.join(eval_output_dir, prefix, "eval_labels.json")
        with open(output_labels_file, "w") as writer:
            json.dump(preds.tolist(), writer, indent=2)
    return results


def load_and_cache_examples(args, task, tokenizer, evaluate=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

    processor = processors[task](task=task)
    output_mode = output_modes[task]
    # Load data features from cache or dataset file

    train_dir = f"{task}" if args.train_suffix is not None and args.train_suffix != "" else task
    test_dir = f"{task}" if args.test_suffix is not None and args.test_suffix != "" else task

    cached_features_file = os.path.join(
        args.test_data_dir if evaluate else args.train_data_dir,
        train_dir if (not evaluate and train_dir is not None) else test_dir,
        "cached_{}_{}_{}".format(
            "test" 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:
        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", args.train_data_dir)
        label_list = processor.get_labels()
        examples, labels = processor.get_test_examples(args.test_data_dir, args.dataset_name) if evaluate else processor.get_train_examples(args.train_data_dir)
        examples = (examples)
        features = convert_examples_to_features(
            examples, tokenizer, max_length=args.max_seq_length, label_list=label_list, output_mode=output_mode,
        )
        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_attention_mask = torch.tensor([f.attention_mask for f in features], dtype=torch.long)
    if args.model_type in ["bert"]:
        all_token_type_ids = torch.tensor([f.token_type_ids for f in features], dtype=torch.long)
    else:
        all_token_type_ids = torch.tensor([1 for f in features], dtype=torch.long)
    if output_mode == "classification":
        # all_labels = torch.tensor([torch.FloatTensor(f.label) for f in features], dtype=torch.long)
        all_labels = torch.from_numpy(labels)
    else:
        raise ValueError("No other `output_mode` for XNLI.")

    dataset = TensorDataset(all_input_ids, all_attention_mask, all_token_type_ids, all_labels)
    return dataset


def main(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
            )
        )

    # 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 = 0 if args.no_cuda else 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)

    # Prepare XNLI task
    if args.dataset_name not in processors:
        raise ValueError("Task not found: %s" % (args.dataset_name))
    processor = processors[args.dataset_name](task=args.dataset_name)
    args.output_mode = output_modes[args.dataset_name]
    label_list = processor.get_labels()
    num_labels = len(label_list)

    # 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

    config = AutoConfig.from_pretrained(
        args.config_name if args.config_name else args.model_name_or_path,
        num_labels=num_labels,
        finetuning_task=args.dataset_name,
        cache_dir=args.cache_dir,
    )
    args.model_type = config.model_type
    tokenizer = AutoTokenizer.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,
    )
    model = AutoModelForSequenceClassification.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.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)

    # Training
    if args.do_train:
        train_dataset = load_and_cache_examples(args, args.dataset_name, tokenizer, evaluate=False)
        global_step, tr_loss = train(args, train_dataset, model, tokenizer)
        logger.info(" global_step = %s, average loss = %s", global_step, tr_loss)

    # Saving best-practices: if you use defaults names for the model, you can reload it using from_pretrained()
    if args.do_train and (args.local_rank == -1 or torch.distributed.get_rank() == 0):
        logger.info("Saving model checkpoint to %s", args.output_dir)
        # Save a trained model, configuration and tokenizer using `save_pretrained()`.
        # They can then be reloaded using `from_pretrained()`
        model_to_save = (
            model.module if hasattr(model, "module") else model
        )  # Take care of distributed/parallel training
        model_to_save.save_pretrained(args.output_dir)
        tokenizer.save_pretrained(args.output_dir)

        # Good practice: save your training arguments together with the trained model
        torch.save(args, os.path.join(args.output_dir, "training_args.bin"))

        # Load a trained model and vocabulary that you have fine-tuned
        model = AutoModelForSequenceClassification.from_pretrained(args.output_dir)
        tokenizer = AutoTokenizer.from_pretrained(args.output_dir)
        model.to(args.device)

    # Evaluation
    results = {}
    if args.do_eval and args.local_rank in [-1, 0]:
        checkpoints = [args.output_dir]
        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 logging
        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 = AutoModelForSequenceClassification.from_pretrained(checkpoint)
            model.to(args.device)
            result = evaluate(args, model, tokenizer, prefix=prefix, save_res=True)
            result = dict((k + "_{}".format(global_step), v) for k, v in result.items())
            results.update(result)

    return results


if __name__ == "__main__":
    parser = argparse.ArgumentParser()

    # Required parameters
    parser.add_argument(
            "--train_data_dir",
            default=INTERMEDIATE_DATA_FOLDER_PATH,
            type=str,
            help="The input data dir. Should contain the .tsv files (or other data files) for the task.",
        )
    parser.add_argument(
        "--test_data_dir",
        default=DATA_FOLDER_PATH,
        type=str,
        help="The input data dir. Should contain the .tsv files (or other data files) for the task.",
    )
    parser.add_argument(
        "--dataset_name",
        default=None,
        type=str,
        help="agnews, nyt_coarse, 20news_coarse, dbpedia, yelp_aspect, yelp, yelp_cost",
    )
    parser.add_argument(
        "--model_name_or_path",
        default=None,  # roberta-base
        type=str,
        required=True,
        help="Path to pretrained model or model identifier from huggingface.co/models",
    )
    parser.add_argument(
        "--train_suffix",
        default=None,
        type=str,
        required=True,
        help="Evaluation language. Also train language if `train_language` is set to None.",
    )
    parser.add_argument(
        "--test_suffix", default=None, type=str, help="Train language if is different of the evaluation language."
    )
    parser.add_argument(
        "--output_dir",
        default=None,
        type=str,
        required=True,
        help="The output directory where the model predictions and checkpoints 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=None,
        type=str,
        help="Where do you want to store the pre-trained models downloaded from s3",
    )
    parser.add_argument(
        "--max_seq_length",
        default=128,
        type=int,
        help="The maximum total input sequence length after tokenization. Sequences longer "
             "than this will be truncated, sequences shorter will be padded.",
    )
    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 test 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("--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(
        "--gradient_accumulation_steps",
        type=int,
        default=1,
        help="Number of updates steps to accumulate before performing a backward/update pass.",
    )
    parser.add_argument("--learning_rate", default=5e-5, type=float, help="The initial learning rate for Adam.")
    parser.add_argument("--weight_decay", default=0.0, type=float, help="Weight decay 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("--logging_steps", type=int, default=500, help="Log every X updates steps.")
    parser.add_argument("--save_steps", type=int, default=500, help="Save checkpoint 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="Avoid using 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(
        "--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("--local_rank", type=int, default=-1, help="For distributed training: local_rank")
    parser.add_argument("--server_ip", type=str, default="", help="For distant debugging.")
    parser.add_argument("--server_port", type=str, default="", help="For distant debugging.")
    parser.add_argument("--granularity", type=str, default="document", help="Select either \"sent\" or \"document\".")
    parser.add_argument("--confidence_threshold", default=0.9, help="Training data confidence threshold.")
    parser.add_argument("--iter", type=int, default=0, help="Iteration # of updating GMM and expanding pseudo-training dataset.")

    args = parser.parse_args()

    main(args)