path: root/transquest/algo/sentence_level/siamesetransquest/run_model.py

import json
import logging
import os
import random
import shutil
from collections import OrderedDict
from typing import List, Dict, Tuple, Iterable, Type, Union, Callable
from zipfile import ZipFile
import requests
import numpy as np
from numpy import ndarray
import transformers
import torch
from sklearn.metrics.pairwise import paired_cosine_distances
from torch import nn, Tensor, device
from torch.optim.optimizer import Optimizer

from torch.utils.data import DataLoader
import torch.multiprocessing as mp
from tqdm.autonotebook import trange
import math
import queue



from . import __version__
from transquest.algo.sentence_level.siamesetransquest.util import http_get, import_from_string, batch_to_device
from transquest.algo.sentence_level.siamesetransquest.evaluation.sentence_evaluator import SentenceEvaluator
from transquest.algo.sentence_level.siamesetransquest.models import Transformer, Pooling
from .evaluation.embedding_similarity_evaluator import EmbeddingSimilarityEvaluator
from .losses.cosine_similarity_loss import CosineSimilarityLoss
from .model_args import SiameseTransQuestArgs
from .readers.input_example import InputExample

logger = logging.getLogger(__name__)


class SiameseTransQuestModel(nn.Sequential):
    """
    Loads or create a SentenceTransformer model, that can be used to map sentences / text to embeddings.

    :param model_name_or_path: If it is a filepath on disc, it loads the model from that path. If it is not a path, it first tries to download a pre-trained SentenceTransformer model. If that fails, tries to construct a model from Huggingface models repository with that name.
    :param device: Device (like 'cuda' / 'cpu') that should be used for computation. If None, checks if a GPU can be used.
    """
    def __init__(self, model_name: str = None, args=None, device: str = None):

        self.args = self._load_model_args(model_name)

        if isinstance(args, dict):
            self.args.update_from_dict(args)
        elif isinstance(args, SiameseTransQuestArgs):
            self.args = args

        if self.args.thread_count:
            torch.set_num_threads(self.args.thread_count)

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

        transformer_model = Transformer(model_name, max_seq_length=80)
        pooling_model = Pooling(transformer_model.get_word_embedding_dimension(), pooling_mode_mean_tokens=True,
                                pooling_mode_cls_token=False,
                                pooling_mode_max_tokens=False)
        modules = [transformer_model, pooling_model]

        if modules is not None and not isinstance(modules, OrderedDict):
            modules = OrderedDict([(str(idx), module) for idx, module in enumerate(modules)])

        super().__init__(modules)
        if device is None:
            device = "cuda" if torch.cuda.is_available() else "cpu"
            logger.info("Use pytorch device: {}".format(device))

        self._target_device = torch.device(device)

    def encode(self, sentences: Union[str, List[str], List[int]],
               batch_size: int = 32,
               show_progress_bar: bool = None,
               output_value: str = 'sentence_embedding',
               convert_to_numpy: bool = True,
               convert_to_tensor: bool = False,
               device: str = None,
               normalize_embeddings: bool = False) -> Union[List[Tensor], ndarray, Tensor]:
        """
        Computes sentence embeddings

        :param sentences: the sentences to embed
        :param batch_size: the batch size used for the computation
        :param show_progress_bar: Output a progress bar when encode sentences
        :param output_value:  Default sentence_embedding, to get sentence embeddings. Can be set to token_embeddings to get wordpiece token embeddings.
        :param convert_to_numpy: If true, the output is a list of numpy vectors. Else, it is a list of pytorch tensors.
        :param convert_to_tensor: If true, you get one large tensor as return. Overwrites any setting from convert_to_numpy
        :param device: Which torch.device to use for the computation
        :param normalize_embeddings: If set to true, returned vectors will have length 1. In that case, the faster dot-product (util.dot_score) instead of cosine similarity can be used.

        :return:
           By default, a list of tensors is returned. If convert_to_tensor, a stacked tensor is returned. If convert_to_numpy, a numpy matrix is returned.
        """
        self.eval()
        if show_progress_bar is None:
            show_progress_bar = (logger.getEffectiveLevel()==logging.INFO or logger.getEffectiveLevel()==logging.DEBUG)

        if convert_to_tensor:
            convert_to_numpy = False

        if output_value == 'token_embeddings':
            convert_to_tensor = False
            convert_to_numpy = False

        input_was_string = False
        if isinstance(sentences, str) or not hasattr(sentences, '__len__'): #Cast an individual sentence to a list with length 1
            sentences = [sentences]
            input_was_string = True

        if device is None:
            device = self._target_device

        self.to(device)

        all_embeddings = []
        length_sorted_idx = np.argsort([-self._text_length(sen) for sen in sentences])
        sentences_sorted = [sentences[idx] for idx in length_sorted_idx]

        for start_index in trange(0, len(sentences), batch_size, desc="Batches", disable=not show_progress_bar):
            sentences_batch = sentences_sorted[start_index:start_index+batch_size]
            features = self.tokenize(sentences_batch)
            features = batch_to_device(features, device)

            with torch.no_grad():
                out_features = self.forward(features)

                if output_value == 'token_embeddings':
                    embeddings = []
                    for token_emb, attention in zip(out_features[output_value], out_features['attention_mask']):
                        last_mask_id = len(attention)-1
                        while last_mask_id > 0 and attention[last_mask_id].item() == 0:
                            last_mask_id -= 1

                        embeddings.append(token_emb[0:last_mask_id+1])
                else:   #Sentence embeddings
                    embeddings = out_features[output_value]
                    embeddings = embeddings.detach()
                    if normalize_embeddings:
                        embeddings = torch.nn.functional.normalize(embeddings, p=2, dim=1)

                    # fixes for #522 and #487 to avoid oom problems on gpu with large datasets
                    if convert_to_numpy:
                        embeddings = embeddings.cpu()

                all_embeddings.extend(embeddings)

        all_embeddings = [all_embeddings[idx] for idx in np.argsort(length_sorted_idx)]

        if convert_to_tensor:
            all_embeddings = torch.stack(all_embeddings)
        elif convert_to_numpy:
            all_embeddings = np.asarray([emb.numpy() for emb in all_embeddings])

        if input_was_string:
            all_embeddings = all_embeddings[0]

        return all_embeddings

    def predict(self, to_predict, verbose=True):
        sentences1 = []
        sentences2 = []

        for text_1, text_2 in to_predict:
            sentences1.append(text_1)
            sentences2.append(text_2)

        embeddings1 = self.encode(sentences1, show_progress_bar=verbose, convert_to_numpy=True)
        embeddings2 = self.encode(sentences2, show_progress_bar=verbose, convert_to_numpy=True)

        cosine_scores = 1 - (paired_cosine_distances(embeddings1, embeddings2))

        return cosine_scores

    def start_multi_process_pool(self, target_devices: List[str] = None):
        """
        Starts multi process to process the encoding with several, independent processes.
        This method is recommended if you want to encode on multiple GPUs. It is advised
        to start only one process per GPU. This method works together with encode_multi_process

        :param target_devices: PyTorch target devices, e.g. cuda:0, cuda:1... If None, all available CUDA devices will be used
        :return: Returns a dict with the target processes, an input queue and and output queue.
        """
        if target_devices is None:
            if torch.cuda.is_available():
                target_devices = ['cuda:{}'.format(i) for i in range(torch.cuda.device_count())]
            else:
                logger.info("CUDA is not available. Start 4 CPU worker")
                target_devices = ['cpu']*4

        logger.info("Start multi-process pool on devices: {}".format(', '.join(map(str, target_devices))))

        ctx = mp.get_context('spawn')
        input_queue = ctx.Queue()
        output_queue = ctx.Queue()
        processes = []

        for cuda_id in target_devices:
            p = ctx.Process(target=SiameseTransQuestModel._encode_multi_process_worker, args=(cuda_id, self, input_queue, output_queue), daemon=True)
            p.start()
            processes.append(p)

        return {'input': input_queue, 'output': output_queue, 'processes': processes}

    @staticmethod
    def stop_multi_process_pool(pool):
        """
        Stops all processes started with start_multi_process_pool
        """
        for p in pool['processes']:
            p.terminate()

        for p in pool['processes']:
            p.join()
            p.close()

        pool['input'].close()
        pool['output'].close()

    def encode_multi_process(self, sentences: List[str], pool: Dict[str, object], batch_size: int = 32, chunk_size: int = None):
        """
        This method allows to run encode() on multiple GPUs. The sentences are chunked into smaller packages
        and sent to individual processes, which encode these on the different GPUs. This method is only suitable
        for encoding large sets of sentences

        :param sentences: List of sentences
        :param pool: A pool of workers started with SentenceTransformer.start_multi_process_pool
        :param batch_size: Encode sentences with batch size
        :param chunk_size: Sentences are chunked and sent to the individual processes. If none, it determine a sensible size.
        :return: Numpy matrix with all embeddings
        """

        if chunk_size is None:
            chunk_size = min(math.ceil(len(sentences) / len(pool["processes"]) / 10), 5000)

        logger.info("Chunk data into packages of size {}".format(chunk_size))

        input_queue = pool['input']
        last_chunk_id = 0
        chunk = []

        for sentence in sentences:
            chunk.append(sentence)
            if len(chunk) >= chunk_size:
                input_queue.put([last_chunk_id, batch_size, chunk])
                last_chunk_id += 1
                chunk = []

        if len(chunk) > 0:
            input_queue.put([last_chunk_id, batch_size, chunk])
            last_chunk_id += 1

        output_queue = pool['output']
        results_list = sorted([output_queue.get() for _ in range(last_chunk_id)], key=lambda x: x[0])
        embeddings = np.concatenate([result[1] for result in results_list])
        return embeddings

    @staticmethod
    def _encode_multi_process_worker(target_device: str, model, input_queue, results_queue):
        """
        Internal working process to encode sentences in multi-process setup
        """
        while True:
            try:
                id, batch_size, sentences = input_queue.get()
                embeddings = model.encode(sentences, device=target_device,  show_progress_bar=False, convert_to_numpy=True, batch_size=batch_size)
                results_queue.put([id, embeddings])
            except queue.Empty:
                break

    def get_max_seq_length(self):
        """
        Returns the maximal sequence length for input the model accepts. Longer inputs will be truncated
        """
        if hasattr(self._first_module(), 'max_seq_length'):
            return self._first_module().max_seq_length

        return None

    def tokenize(self, text: str):
        """
        Tokenizes the text
        """
        return self._first_module().tokenize(text)

    def get_sentence_features(self, *features):
        return self._first_module().get_sentence_features(*features)

    def get_sentence_embedding_dimension(self):
        for mod in reversed(self._modules.values()):
            sent_embedding_dim_method = getattr(mod, "get_sentence_embedding_dimension", None)
            if callable(sent_embedding_dim_method):
                return sent_embedding_dim_method()
        return None

    def _first_module(self):
        """Returns the first module of this sequential embedder"""
        return self._modules[next(iter(self._modules))]

    def _last_module(self):
        """Returns the last module of this sequential embedder"""
        return self._modules[next(reversed(self._modules))]

    def save(self, path):
        """
        Saves all elements for this seq. sentence embedder into different sub-folders
        """
        if path is None:
            return

        os.makedirs(path, exist_ok=True)

        logger.info("Save model to {}".format(path))
        contained_modules = []

        for idx, name in enumerate(self._modules):
            module = self._modules[name]
            # model_path = os.path.join(path, str(idx)+"_"+type(module).__name__)
            os.makedirs(path, exist_ok=True)
            module.save(path)
            contained_modules.append({'idx': idx, 'name': name, 'path': os.path.basename(path), 'type': type(module).__module__})

        with open(os.path.join(path, 'modules.json'), 'w') as fOut:
            json.dump(contained_modules, fOut, indent=2)

        with open(os.path.join(path, 'siamese_config.json'), 'w') as fOut:
            json.dump({'__version__': __version__}, fOut, indent=2)

    def smart_batching_collate(self, batch):
        """
        Transforms a batch from a SmartBatchingDataset to a batch of tensors for the model
        Here, batch is a list of tuples: [(tokens, label), ...]

        :param batch:
            a batch from a SmartBatchingDataset
        :return:
            a batch of tensors for the model
        """
        num_texts = len(batch[0].texts)
        texts = [[] for _ in range(num_texts)]
        labels = []

        for example in batch:
            for idx, text in enumerate(example.texts):
                texts[idx].append(text)

            labels.append(example.label)

        labels = torch.tensor(labels).to(self._target_device)

        sentence_features = []
        for idx in range(num_texts):
            tokenized = self.tokenize(texts[idx])
            batch_to_device(tokenized, self._target_device)
            sentence_features.append(tokenized)

        return sentence_features, labels

    def _text_length(self, text: Union[List[int], List[List[int]]]):
        """
        Help function to get the length for the input text. Text can be either
        a list of ints (which means a single text as input), or a tuple of list of ints
        (representing several text inputs to the model).
        """

        if isinstance(text, dict):              #{key: value} case
            return len(next(iter(text.values())))
        elif not hasattr(text, '__len__'):      #Object has no len() method
            return 1
        elif len(text) == 0 or isinstance(text[0], int):    #Empty string or list of ints
            return len(text)
        else:
            return sum([len(t) for t in text])      #Sum of length of individual strings

    def train_model(self, train_df, eval_df, args=None, output_dir=None, verbose=True):

        train_samples = []
        for index, row in train_df.iterrows():
            score = float(row["labels"])
            inp_example = InputExample(texts=[row['text_a'], row['text_b']], label=score)
            train_samples.append(inp_example)

        eval_samples = []
        for index, row in eval_df.iterrows():
            score = float(row["labels"])
            inp_example = InputExample(texts=[row['text_a'], row['text_b']], label=score)
            eval_samples.append(inp_example)

        train_dataloader = DataLoader(train_samples, shuffle=True,
                                      batch_size=self.args.train_batch_size)
        train_loss = CosineSimilarityLoss(model=self)

        evaluator = EmbeddingSimilarityEvaluator.from_input_examples(eval_samples, name='eval')
        warmup_steps = math.ceil(len(train_dataloader) * self.args.num_train_epochs * 0.1)

        self.fit(train_objectives=[(train_dataloader, train_loss)],
                  evaluator=evaluator,
                  epochs=self.args.num_train_epochs,
                  evaluation_steps=self.args.evaluate_during_training_steps,
                  optimizer_params={'lr': self.args.learning_rate,
                                    'eps': self.args.adam_epsilon,
                                    'correct_bias': False},
                  warmup_steps=warmup_steps,
                  weight_decay=self.args.weight_decay,
                  max_grad_norm=self.args.max_grad_norm,
                  output_path=self.args.best_model_dir)


    def fit(self,
            train_objectives: Iterable[Tuple[DataLoader, nn.Module]],
            evaluator: SentenceEvaluator = None,
            epochs: int = 1,
            steps_per_epoch = None,
            scheduler: str = 'WarmupLinear',
            warmup_steps: int = 10000,
            optimizer_class: Type[Optimizer] = transformers.AdamW,
            optimizer_params : Dict[str, object]= {'lr': 2e-5},
            weight_decay: float = 0.01,
            evaluation_steps: int = 0,
            output_path: str = None,
            save_best_model: bool = True,
            max_grad_norm: float = 1,
            use_amp: bool = False,
            callback: Callable[[float, int, int], None] = None,
            show_progress_bar: bool = True
            ):
        """
        Train the model with the given training objective
        Each training objective is sampled in turn for one batch.
        We sample only as many batches from each objective as there are in the smallest one
        to make sure of equal training with each dataset.

        :param train_objectives: Tuples of (DataLoader, LossFunction). Pass more than one for multi-task learning
        :param evaluator: An evaluator (sentence_transformers.evaluation) evaluates the model performance during training on held-out dev data. It is used to determine the best model that is saved to disc.
        :param epochs: Number of epochs for training
        :param steps_per_epoch: Number of training steps per epoch. If set to None (default), one epoch is equal the DataLoader size from train_objectives.
        :param scheduler: Learning rate scheduler. Available schedulers: constantlr, warmupconstant, warmuplinear, warmupcosine, warmupcosinewithhardrestarts
        :param warmup_steps: Behavior depends on the scheduler. For WarmupLinear (default), the learning rate is increased from o up to the maximal learning rate. After these many training steps, the learning rate is decreased linearly back to zero.
        :param optimizer_class: Optimizer
        :param optimizer_params: Optimizer parameters
        :param weight_decay: Weight decay for model parameters
        :param evaluation_steps: If > 0, evaluate the model using evaluator after each number of training steps
        :param output_path: Storage path for the model and evaluation files
        :param save_best_model: If true, the best model (according to evaluator) is stored at output_path
        :param max_grad_norm: Used for gradient normalization.
        :param use_amp: Use Automatic Mixed Precision (AMP). Only for Pytorch >= 1.6.0
        :param callback: Callback function that is invoked after each evaluation.
                It must accept the following three parameters in this order:
                `score`, `epoch`, `steps`
        :param show_progress_bar: If True, output a tqdm progress bar
        """

        if use_amp:
            from torch.cuda.amp import autocast
            scaler = torch.cuda.amp.GradScaler()

        self.to(self._target_device)

        if output_path is not None:
            os.makedirs(output_path, exist_ok=True)

        dataloaders = [dataloader for dataloader, _ in train_objectives]

        # Use smart batching
        for dataloader in dataloaders:
            dataloader.collate_fn = self.smart_batching_collate

        loss_models = [loss for _, loss in train_objectives]
        for loss_model in loss_models:
            loss_model.to(self._target_device)

        self.best_score = -9999999

        if steps_per_epoch is None or steps_per_epoch == 0:
            steps_per_epoch = min([len(dataloader) for dataloader in dataloaders])

        num_train_steps = int(steps_per_epoch * epochs)

        # Prepare optimizers
        optimizers = []
        schedulers = []
        for loss_model in loss_models:
            param_optimizer = list(loss_model.named_parameters())

            no_decay = ['bias', 'LayerNorm.bias', 'LayerNorm.weight']
            optimizer_grouped_parameters = [
                {'params': [p for n, p in param_optimizer if not any(nd in n for nd in no_decay)], 'weight_decay': weight_decay},
                {'params': [p for n, p in param_optimizer if any(nd in n for nd in no_decay)], 'weight_decay': 0.0}
            ]

            optimizer = optimizer_class(optimizer_grouped_parameters, **optimizer_params)
            scheduler_obj = self._get_scheduler(optimizer, scheduler=scheduler, warmup_steps=warmup_steps, t_total=num_train_steps)

            optimizers.append(optimizer)
            schedulers.append(scheduler_obj)


        global_step = 0
        data_iterators = [iter(dataloader) for dataloader in dataloaders]

        num_train_objectives = len(train_objectives)

        skip_scheduler = False
        for epoch in trange(epochs, desc="Epoch", disable=not show_progress_bar):
            training_steps = 0

            for loss_model in loss_models:
                loss_model.zero_grad()
                loss_model.train()

            for _ in trange(steps_per_epoch, desc="Iteration", smoothing=0.05, disable=not show_progress_bar):
                for train_idx in range(num_train_objectives):
                    loss_model = loss_models[train_idx]
                    optimizer = optimizers[train_idx]
                    scheduler = schedulers[train_idx]
                    data_iterator = data_iterators[train_idx]

                    try:
                        data = next(data_iterator)
                    except StopIteration:
                        data_iterator = iter(dataloaders[train_idx])
                        data_iterators[train_idx] = data_iterator
                        data = next(data_iterator)


                    features, labels = data


                    if use_amp:
                        with autocast():
                            loss_value = loss_model(features, labels)

                        scale_before_step = scaler.get_scale()
                        scaler.scale(loss_value).backward()
                        scaler.unscale_(optimizer)
                        torch.nn.utils.clip_grad_norm_(loss_model.parameters(), max_grad_norm)
                        scaler.step(optimizer)
                        scaler.update()

                        skip_scheduler = scaler.get_scale() != scale_before_step
                    else:
                        loss_value = loss_model(features, labels)
                        loss_value.backward()
                        torch.nn.utils.clip_grad_norm_(loss_model.parameters(), max_grad_norm)
                        optimizer.step()

                    optimizer.zero_grad()

                    if not skip_scheduler:
                        scheduler.step()

                training_steps += 1
                global_step += 1

                if evaluation_steps > 0 and training_steps % evaluation_steps == 0:
                    self._eval_during_training(evaluator, output_path, save_best_model, epoch,
                                               training_steps, callback)
                    for loss_model in loss_models:
                        loss_model.zero_grad()
                        loss_model.train()

            self._eval_during_training(evaluator, output_path, save_best_model, epoch, -1, callback)

        if evaluator is None and output_path is not None:   #No evaluator, but output path: save final model version
            self.save(output_path)

    def evaluate(self, evaluator: SentenceEvaluator, output_path: str = None, verbose: bool = True):
        """
        Evaluate the model

        :param evaluator:
            the evaluator
        :param verbose:
            print the results
        :param output_path:
            the evaluator can write the results to this path
        """
        if output_path is not None:
            os.makedirs(output_path, exist_ok=True)
        return evaluator(self, output_path, verbose)

    def _eval_during_training(self, evaluator, output_path, save_best_model, epoch, steps, callback):
        """Runs evaluation during the training"""
        if evaluator is not None:
            score = evaluator(self, output_path=output_path, epoch=epoch, steps=steps)
            if callback is not None:
                callback(score, epoch, steps)
            if score > self.best_score:
                self.best_score = score
                if save_best_model:
                    self.save(output_path)

    @staticmethod
    def _get_scheduler(optimizer, scheduler: str, warmup_steps: int, t_total: int):
        """
        Returns the correct learning rate scheduler. Available scheduler: constantlr, warmupconstant, warmuplinear, warmupcosine, warmupcosinewithhardrestarts
        """
        scheduler = scheduler.lower()
        if scheduler == 'constantlr':
            return transformers.get_constant_schedule(optimizer)
        elif scheduler == 'warmupconstant':
            return transformers.get_constant_schedule_with_warmup(optimizer, num_warmup_steps=warmup_steps)
        elif scheduler == 'warmuplinear':
            return transformers.get_linear_schedule_with_warmup(optimizer, num_warmup_steps=warmup_steps, num_training_steps=t_total)
        elif scheduler == 'warmupcosine':
            return transformers.get_cosine_schedule_with_warmup(optimizer, num_warmup_steps=warmup_steps, num_training_steps=t_total)
        elif scheduler == 'warmupcosinewithhardrestarts':
            return transformers.get_cosine_with_hard_restarts_schedule_with_warmup(optimizer, num_warmup_steps=warmup_steps, num_training_steps=t_total)
        else:
            raise ValueError("Unknown scheduler {}".format(scheduler))

    @property
    def device(self) -> device:
        """
        Get torch.device from module, assuming that the whole module has one device.
        """
        try:
            return next(self.parameters()).device
        except StopIteration:
            # For nn.DataParallel compatibility in PyTorch 1.5

            def find_tensor_attributes(module: nn.Module) -> List[Tuple[str, Tensor]]:
                tuples = [(k, v) for k, v in module.__dict__.items() if torch.is_tensor(v)]
                return tuples

            gen = self._named_members(get_members_fn=find_tensor_attributes)
            first_tuple = next(gen)
            return first_tuple[1].device

    def save_model_args(self, output_dir):
        os.makedirs(output_dir, exist_ok=True)
        self.args.save(output_dir)

    def _load_model_args(self, input_dir):
        args = SiameseTransQuestArgs()
        args.load(input_dir)
        return args

    @property
    def tokenizer(self):
        """
        Property to get the tokenizer that is used by this model
        """
        return self._first_module().tokenizer

    @tokenizer.setter
    def tokenizer(self, value):
        """
        Property to set the tokenizer that is should used by this model
        """
        self._first_module().tokenizer = value

    @property
    def max_seq_length(self):
        """
        Property to get the maximal input sequence length for the model. Longer inputs will be truncated.
        """
        return self._first_module().max_seq_length

    @max_seq_length.setter
    def max_seq_length(self, value):
        """
        Property to set the maximal input sequence length for the model. Longer inputs will be truncated.
        """
        self._first_module().max_seq_length = value