finetuner.py

#!/usr/bin/env python
import copy
import logging
import os
import sys
from copy import deepcopy
from itertools import chain
from typing import Union

import datasets
import evaluate
import torch
import transformers
from transformers import (
    Trainer,
    default_data_collator,
    set_seed,
)
from transformers.trainer_utils import get_last_checkpoint

try:
    from transformers.utils import send_example_telemetry
except ImportError:
    send_example_telemetry = None

from lmflow.args import DatasetArguments, FinetunerArguments, ModelArguments
from lmflow.datasets.dataset import Dataset
from lmflow.models.hf_decoder_model import HFDecoderModel

# from lmflow.models.hf_encoder_decoder_model import HFEncoderDecoderModel
from lmflow.models.hf_text_regression_model import HFTextRegressionModel
from lmflow.optim import create_customized_optimizer
from lmflow.pipeline.base_tuner import BaseTuner
from lmflow.pipeline.utils.lisa_trainer import DynamicLayerActivationCallback
from lmflow.utils.versioning import is_package_version_at_least

logger = logging.getLogger(__name__)


class Finetuner(BaseTuner):
    """
    Initializes the `Finetuner` class with given arguments.

    Parameters
    ------------
    model_args : ModelArguments object.
        Contains the arguments required to load the model.

    data_args : DatasetArguments object.
        Contains the arguments required to load the dataset.

    finetuner_args : FinetunerArguments object.
        Contains the arguments required to perform finetuning.

    args : Optional.
        Positional arguments.

    kwargs : Optional.
        Keyword arguments.

    """

    def __init__(
        self,
        model_args: ModelArguments,
        data_args: DatasetArguments,
        finetuner_args: FinetunerArguments,
        *args,
        **kwargs,
    ):
        self.model_args = model_args
        self.data_args = data_args
        self.finetuner_args = finetuner_args

        # Sending telemetry. Tracking the example usage helps us better
        # allocate resources to maintain them. The information sent is the one
        # passed as arguments along with your Python/PyTorch versions.
        if send_example_telemetry is not None:
            send_example_telemetry("run_clm", model_args, data_args)

        # 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 = finetuner_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: {finetuner_args.local_rank},"
            f" device: {finetuner_args.device},"
            f" n_gpu: {finetuner_args.n_gpu},"
            f"distributed training: {bool(finetuner_args.local_rank != -1)},"
            f" 16-bits training: {finetuner_args.fp16}"
        )
        logger.info(f"Training/evaluation parameters {finetuner_args}")

        # Detecting last checkpoint.
        last_checkpoint = None
        if (
            os.path.isdir(finetuner_args.output_dir)
            and finetuner_args.do_train
            and not finetuner_args.overwrite_output_dir
        ):
            last_checkpoint = get_last_checkpoint(finetuner_args.output_dir)
            if last_checkpoint is None and len(os.listdir(finetuner_args.output_dir)) > 0:
                raise ValueError(
                    f"Output directory ({finetuner_args.output_dir}) already"
                    " exists and is not empty. "
                    "Use --overwrite_output_dir to overcome."
                )
            elif last_checkpoint is not None and finetuner_args.resume_from_checkpoint is None:
                logger.info(
                    f"Checkpoint detected, resuming training at"
                    f" {last_checkpoint}. To avoid this behavior, change"
                    " the `--output_dir` or add `--overwrite_output_dir` to"
                    " train from scratch."
                )
        self.last_checkpoint = last_checkpoint

        # Set seed before initializing model.
        set_seed(finetuner_args.seed)

    def group_text(self, tokenized_datasets, model_max_length):
        """
        Groups texts together to form blocks of maximum length `model_max_length` and returns the processed data as
        a dictionary.
        """
        data_args = self.data_args
        finetuner_args = self.finetuner_args

        if data_args.block_size is None:
            block_size = model_max_length
            if block_size > 1024:
                logger.warning(
                    "The chosen tokenizer supports a `model_max_length` that is"
                    " longer than the default `block_size` value"
                    " of 1024. If you would like to use a longer `block_size`"
                    " up to `tokenizer.model_max_length` you can override this "
                    " default with `--block_size xxx`."
                )
                block_size = 1024
        else:
            if data_args.block_size > model_max_length:
                if self.model_args.truncate_to_model_max_length:
                    logger.warning(
                        f"The block_size passed ({data_args.block_size}) is larger"
                        f" than the maximum length for the model"
                        f"({model_max_length})."
                        f" Using block_size={model_max_length}."
                        f"If you would like to use a longer 'block_size' that is"
                        f" longer than the maximum length supported by the model,"
                        f" you can override this behavior with"
                        f"default with `--truncate_to_model_max_length False`."
                    )
                    block_size = model_max_length
                else:
                    logger.warning(
                        f"The block_size passed ({data_args.block_size}) is larger"
                        f"than the maximum length for the model"
                        f"({model_max_length})."
                        f"Using block_size={data_args.block_size}."
                    )
                    block_size = data_args.block_size
            else:
                block_size = data_args.block_size

        # Main data processing function that will concatenate all texts from
        # our dataset and generate chunks of block_size.
        def group_texts(examples):
            # Concatenate all texts.
            concatenated_examples = {k: list(chain(*examples[k])) for k in examples.keys()}
            total_length = len(concatenated_examples[list(examples.keys())[0]])
            # We drop the small remainder, we could add padding if the model
            # supported it instead of this drop, you can customize this part to
            # your needs.
            total_length = (total_length // block_size) * block_size
            # Split by chunks of max_len.
            result = {
                k: [t[i : i + block_size] for i in range(0, total_length, block_size)]
                for k, t in concatenated_examples.items()
            }
            return result

        # Note that with `batched=True`, this map processes 1,000 texts
        # together, so group_texts throws away a remainder for each of those
        # groups of 1,000 texts. You can adjust that batch_size here but a
        # higher value might be slower to preprocess.
        #
        # To speed up this part, we use multiprocessing. See the documentation
        # of the map method for more information:
        # https://huggingface.co/docs/datasets/package_reference/main_classes.html#datasets.Dataset.map
        with finetuner_args.main_process_first(desc="grouping texts together"):
            group_batch_size = data_args.group_texts_batch_size
            if data_args.disable_group_texts:
                group_batch_size = 1
            if not data_args.streaming:
                lm_datasets = tokenized_datasets.map(
                    group_texts,
                    batched=True,
                    batch_size=group_batch_size,
                    num_proc=data_args.preprocessing_num_workers,
                    load_from_cache_file=not data_args.overwrite_cache,
                    desc=f"Grouping texts in chunks of {block_size}",
                )
            else:
                lm_datasets = tokenized_datasets.map(
                    group_texts,
                    batched=True,
                    batch_size=group_batch_size,
                )

        return lm_datasets

    def tune(
        self,
        model: Union[HFDecoderModel, HFTextRegressionModel],
        dataset: Dataset,
        transform_dataset_in_place=True,
        data_collator=None,
    ):
        """
        Perform tuning for a model

        Parameters
        ------------
        model : TunableModel object.
            TunableModel to perform tuning.

        dataset:
            dataset to train model.

        """
        model_args = self.model_args
        data_args = self.data_args
        finetuner_args = self.finetuner_args
        if not transform_dataset_in_place:
            dataset = copy.deepcopy(dataset)

        # Tokenization and text grouping must be done in the main process
        if dataset.backend == "custom_multi_modal":
            dataset.backend_dataset.register_tokenizer(model.tokenizer, model.image_processor)
            lm_dataset = dataset
        else:
            with finetuner_args.main_process_first(desc="dataset map tokenization"):
                tokenized_dataset = model.tokenize(dataset)
                if data_args.disable_group_texts:
                    lm_dataset = tokenized_dataset
                else:
                    lm_dataset = self.group_text(
                        tokenized_dataset,
                        model_max_length=model.get_max_length(),
                    )

        train_dataset = lm_dataset.get_backend_dataset()

        if data_args.calculate_dataset_stats:
            total_tokens = 0
            total_target_tokens = 0
            pad_token_id = model.get_tokenizer().pad_token_id
            logger.warning("Calculating dataset stats...")
            import time

            start_time = time.time()
            for datapoint in train_dataset:
                total_tokens += len([label for label in datapoint["input_ids"] if label != pad_token_id])
                total_target_tokens += len([label for label in datapoint["labels"] if label != -100])
            logger.warning(
                "Dataset stats:\n\n"
                f"Total tokens: {total_tokens}\n"
                f"Total target tokens: {total_target_tokens}\n"
                f"Total samples: {len(train_dataset)}\n"
                f"Average tokens per sample: {total_tokens / len(train_dataset)}\n"
                f"Average target tokens per sample: {total_target_tokens / len(train_dataset)}\n"
            )
            logger.warning("Calculating data stats took %s seconds", time.time() - start_time)
        else:
            logger.warning(f"Number of train samples: {len(train_dataset)}")

        if finetuner_args.do_eval:
            eval_dataset_args = deepcopy(data_args)
            eval_dataset_args.dataset_path = finetuner_args.eval_dataset_path
            eval_dataset = Dataset(eval_dataset_args)
            with finetuner_args.main_process_first(desc="dataset map tokenization"):
                tokenized_dataset = model.tokenize(eval_dataset)
                if data_args.disable_group_texts:
                    lm_dataset = tokenized_dataset
                else:
                    lm_dataset = self.group_text(
                        tokenized_dataset,
                        model_max_length=model.get_max_length(),
                    )
            eval_dataset = lm_dataset.get_backend_dataset()
            logger.info(f"Number of eval samples: {len(eval_dataset)}")

            def preprocess_logits_for_metrics(logits, labels):
                if isinstance(logits, tuple):
                    # Depending on the model and config, logits may contain extra tensors,
                    # like past_key_values, but logits always come first
                    logits = logits[0]
                return logits.argmax(dim=-1)

            metric = evaluate.load("accuracy")

            def compute_metrics(eval_preds):
                preds, labels = eval_preds
                # preds have the same shape as the labels, after the argmax(-1) has been calculated
                # by preprocess_logits_for_metrics but we need to shift the labels
                labels = labels[:, 1:].reshape(-1)
                preds = preds[:, :-1].reshape(-1)
                return metric.compute(predictions=preds, references=labels)

        if finetuner_args.do_train:
            if data_args.max_train_samples is not None:
                max_train_samples = min(len(train_dataset), data_args.max_train_samples)
                train_dataset = train_dataset.select(range(max_train_samples))

        # Initialize our Trainer
        training_args = finetuner_args
        FinetuningTrainer = Trainer
        trainer_callbacks = []

        if data_collator is None:
            data_collator = default_data_collator

        if training_args.use_customized_optim:
            BaseTrainer = FinetuningTrainer
            FinetuningTrainer = create_customized_optimizer(BaseTrainer, model_args)

        if training_args.use_lisa:
            dynamic_layer_activation_callback = DynamicLayerActivationCallback(
                n_layers=training_args.lisa_activated_layers,  # Number of layers to activate
                interval_steps=training_args.lisa_interval_steps,  # Step interval to update active layers
                model=model.get_backend_model(),
                lisa_layers_attribute=training_args.lisa_layers_attribute,  # Attribute to access layers of the model
            )
            trainer_callbacks.append(dynamic_layer_activation_callback)

        trainer_kwargs = {
            "model": model.get_backend_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,
            "data_collator": data_collator,
            "compute_metrics": compute_metrics if training_args.do_eval else None,
            "preprocess_logits_for_metrics": preprocess_logits_for_metrics if training_args.do_eval else None,
            "callbacks": trainer_callbacks,
        }
        if is_package_version_at_least("transformers", "4.46.0"):
            # https://github.com/huggingface/transformers/pull/32385
            trainer_kwargs["processing_class"] = model.get_tokenizer()
        else:
            trainer_kwargs["tokenizer"] = model.get_tokenizer()
        trainer = FinetuningTrainer(**trainer_kwargs)

        # Training
        if training_args.do_train:
            checkpoint = None
            last_checkpoint = self.last_checkpoint
            if training_args.resume_from_checkpoint is not None:
                # load from lora checkpoint is also supported
                checkpoint = training_args.resume_from_checkpoint
            elif last_checkpoint is not None:
                checkpoint = last_checkpoint
            train_result = trainer.train(resume_from_checkpoint=checkpoint)

            if not model_args.use_lora:
                trainer.save_model()  # Saves the tokenizer too for easy upload
            else:
                if model_args.save_aggregated_lora:
                    model.merge_lora_weights()
                model.save(finetuner_args.output_dir, model_args.save_aggregated_lora)
            # save language_projection for multi-modal model;
            if self.finetuner_args.save_language_projection:
                language_projection_state = trainer.model.language_projection.state_dict()
                torch.save(
                    os.path.join(self.finetuner_args.output_dir, "language_projection.pth"), language_projection_state
                )
            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.log_metrics("train", metrics)
            trainer.save_metrics("train", metrics)
            trainer.save_state()

        kwargs = {"finetuned_from": model_args.model_name_or_path, "tasks": "text-generation"}
        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

        if training_args.push_to_hub:
            trainer.push_to_hub(**kwargs)
        else:
            trainer.create_model_card(**kwargs)

        return model