From 775d0d24ed1f7a8d594caca68a4c2ba833237120 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E7=89=9B=E5=86=A0=E5=8D=9A?= Date: Fri, 17 Feb 2023 23:43:44 +0800 Subject: [PATCH 1/4] add bert squad code --- nlp/question_answering/bert/pytorch/README.md | 35 + .../bert/pytorch/requirements.txt | 4 + nlp/question_answering/bert/pytorch/run.sh | 11 + .../bert/pytorch/run_dist.sh | 12 + nlp/question_answering/bert/pytorch/run_qa.py | 681 ++++++++++++++++++ .../bert/pytorch/squad_download.py | 153 ++++ .../bert/pytorch/trainer_qa.py | 136 ++++ .../bert/pytorch/utils_qa.py | 443 ++++++++++++ 8 files changed, 1475 insertions(+) create mode 100644 nlp/question_answering/bert/pytorch/README.md create mode 100644 nlp/question_answering/bert/pytorch/requirements.txt create mode 100644 nlp/question_answering/bert/pytorch/run.sh create mode 100644 nlp/question_answering/bert/pytorch/run_dist.sh create mode 100644 nlp/question_answering/bert/pytorch/run_qa.py create mode 100644 nlp/question_answering/bert/pytorch/squad_download.py create mode 100644 nlp/question_answering/bert/pytorch/trainer_qa.py create mode 100644 nlp/question_answering/bert/pytorch/utils_qa.py diff --git a/nlp/question_answering/bert/pytorch/README.md b/nlp/question_answering/bert/pytorch/README.md new file mode 100644 index 00000000..95183adc --- /dev/null +++ b/nlp/question_answering/bert/pytorch/README.md @@ -0,0 +1,35 @@ +# Bert-base squad + +## Model description + +Bert-base squad task Fine-tuning + +## Step 1: Installing packages + +``` shell +cd /nlp/querstion_answering/bert/pytorch +pip3 install -r requirements.txt +``` + +## Step 2: Training + +### On single GPU + +``` shell +bash run.sh +``` + +### Multiple GPUs on one machine + +```shell +bash run_dist.sh +``` +## Results on BI-V100 + +| GPUs | Samples/s | Loss | +|------|-----------|--------| +| 1x1 | 29.86 | 0.9861 | +| 1x8 | 178.906 | 0.9804 | + +## Reference +https://github.com/huggingface/ \ No newline at end of file diff --git a/nlp/question_answering/bert/pytorch/requirements.txt b/nlp/question_answering/bert/pytorch/requirements.txt new file mode 100644 index 00000000..c66e961c --- /dev/null +++ b/nlp/question_answering/bert/pytorch/requirements.txt @@ -0,0 +1,4 @@ +accelerate >= 0.12.0 +datasets >= 1.8.0 +evaluate +git+https://github.com/huggingface/transformers \ No newline at end of file diff --git a/nlp/question_answering/bert/pytorch/run.sh b/nlp/question_answering/bert/pytorch/run.sh new file mode 100644 index 00000000..cf87d4c6 --- /dev/null +++ b/nlp/question_answering/bert/pytorch/run.sh @@ -0,0 +1,11 @@ +python3 run_qa.py \ + --model_name_or_path "/home/data/bert/bert-base-uncased-pt" \ + --dataset_name squad \ + --do_train \ + --do_eval \ + --per_device_train_batch_size 12 \ + --learning_rate 3e-5 \ + --num_train_epochs 2 \ + --max_seq_length 384 \ + --doc_stride 128 \ + --output_dir /tmp/debug_squad/ \ No newline at end of file diff --git a/nlp/question_answering/bert/pytorch/run_dist.sh b/nlp/question_answering/bert/pytorch/run_dist.sh new file mode 100644 index 00000000..efc59ad4 --- /dev/null +++ b/nlp/question_answering/bert/pytorch/run_dist.sh @@ -0,0 +1,12 @@ +python3 -m torch.distributed.launch --nproc_per_node=8 --master_port 12333 \ + run_qa.py \ + --model_name_or_path "/home/data/bert/bert-base-uncased-pt" \ + --dataset_name squad \ + --do_train \ + --do_eval \ + --per_device_train_batch_size 12 \ + --learning_rate 3e-5 \ + --num_train_epochs 2 \ + --max_seq_length 384 \ + --doc_stride 128 \ + --output_dir /tmp/debug_squad/ \ No newline at end of file diff --git a/nlp/question_answering/bert/pytorch/run_qa.py b/nlp/question_answering/bert/pytorch/run_qa.py new file mode 100644 index 00000000..1c7a734d --- /dev/null +++ b/nlp/question_answering/bert/pytorch/run_qa.py @@ -0,0 +1,681 @@ +#!/usr/bin/env python +# coding=utf-8 +# Copyright 2020 The HuggingFace Team All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" +Fine-tuning the library models for question answering using a slightly adapted version of the 🤗 Trainer. +""" +# You can also adapt this script on your own question answering task. Pointers for this are left as comments. + +import logging +import os +import sys +from dataclasses import dataclass, field +from typing import Optional + +import datasets +import evaluate +from datasets import load_dataset +from trainer_qa import QuestionAnsweringTrainer +from utils_qa import postprocess_qa_predictions + +import transformers +from transformers import ( + AutoConfig, + AutoModelForQuestionAnswering, + AutoTokenizer, + DataCollatorWithPadding, + EvalPrediction, + HfArgumentParser, + PreTrainedTokenizerFast, + TrainingArguments, + default_data_collator, + set_seed, +) +from transformers.trainer_utils import get_last_checkpoint +from transformers.utils import check_min_version, send_example_telemetry +from transformers.utils.versions import require_version + +# Will error if the minimal version of Transformers is not installed. Remove at your own risks. +check_min_version("4.27.0.dev0") + +require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/question-answering/requirements.txt") + +logger = logging.getLogger(__name__) + + +@dataclass +class ModelArguments: + """ + Arguments pertaining to which model/config/tokenizer we are going to fine-tune from. + """ + + model_name_or_path: str = field( + metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"} + ) + config_name: Optional[str] = field( + default=None, metadata={"help": "Pretrained config name or path if not the same as model_name"} + ) + tokenizer_name: Optional[str] = field( + default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"} + ) + cache_dir: Optional[str] = field( + default=None, + metadata={"help": "Path to directory to store the pretrained models downloaded from huggingface.co"}, + ) + model_revision: str = field( + default="main", + metadata={"help": "The specific model version to use (can be a branch name, tag name or commit id)."}, + ) + use_auth_token: bool = field( + default=False, + metadata={ + "help": ( + "Will use the token generated when running `huggingface-cli login` (necessary to use this script " + "with private models)." + ) + }, + ) + + +@dataclass +class DataTrainingArguments: + """ + Arguments pertaining to what data we are going to input our model for training and eval. + """ + + dataset_name: Optional[str] = field( + default=None, metadata={"help": "The name of the dataset to use (via the datasets library)."} + ) + dataset_config_name: Optional[str] = field( + default=None, metadata={"help": "The configuration name of the dataset to use (via the datasets library)."} + ) + train_file: Optional[str] = field(default=None, metadata={"help": "The input training data file (a text file)."}) + validation_file: Optional[str] = field( + default=None, + metadata={"help": "An optional input evaluation data file to evaluate the perplexity on (a text file)."}, + ) + test_file: Optional[str] = field( + default=None, + metadata={"help": "An optional input test data file to evaluate the perplexity on (a text file)."}, + ) + overwrite_cache: bool = field( + default=False, metadata={"help": "Overwrite the cached training and evaluation sets"} + ) + preprocessing_num_workers: Optional[int] = field( + default=None, + metadata={"help": "The number of processes to use for the preprocessing."}, + ) + max_seq_length: int = field( + default=384, + metadata={ + "help": ( + "The maximum total input sequence length after tokenization. Sequences longer " + "than this will be truncated, sequences shorter will be padded." + ) + }, + ) + pad_to_max_length: bool = field( + default=True, + metadata={ + "help": ( + "Whether to pad all samples to `max_seq_length`. If False, will pad the samples dynamically when" + " batching to the maximum length in the batch (which can be faster on GPU but will be slower on TPU)." + ) + }, + ) + max_train_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of training examples to this " + "value if set." + ) + }, + ) + max_eval_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of evaluation examples to this " + "value if set." + ) + }, + ) + max_predict_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of prediction examples to this " + "value if set." + ) + }, + ) + version_2_with_negative: bool = field( + default=False, metadata={"help": "If true, some of the examples do not have an answer."} + ) + null_score_diff_threshold: float = field( + default=0.0, + metadata={ + "help": ( + "The threshold used to select the null answer: if the best answer has a score that is less than " + "the score of the null answer minus this threshold, the null answer is selected for this example. " + "Only useful when `version_2_with_negative=True`." + ) + }, + ) + doc_stride: int = field( + default=128, + metadata={"help": "When splitting up a long document into chunks, how much stride to take between chunks."}, + ) + n_best_size: int = field( + default=20, + metadata={"help": "The total number of n-best predictions to generate when looking for an answer."}, + ) + max_answer_length: int = field( + default=30, + metadata={ + "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." + ) + }, + ) + + def __post_init__(self): + if ( + self.dataset_name is None + and self.train_file is None + and self.validation_file is None + and self.test_file is None + ): + raise ValueError("Need either a dataset name or a training/validation file/test_file.") + else: + if self.train_file is not None: + extension = self.train_file.split(".")[-1] + assert extension in ["csv", "json"], "`train_file` should be a csv or a json file." + if self.validation_file is not None: + extension = self.validation_file.split(".")[-1] + assert extension in ["csv", "json"], "`validation_file` should be a csv or a json file." + if self.test_file is not None: + extension = self.test_file.split(".")[-1] + assert extension in ["csv", "json"], "`test_file` should be a csv or a json file." + + +def main(): + # See all possible arguments in src/transformers/training_args.py + # or by passing the --help flag to this script. + # We now keep distinct sets of args, for a cleaner separation of concerns. + + parser = HfArgumentParser((ModelArguments, DataTrainingArguments, TrainingArguments)) + if len(sys.argv) == 2 and sys.argv[1].endswith(".json"): + # If we pass only one argument to the script and it's the path to a json file, + # let's parse it to get our arguments. + model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1])) + else: + model_args, data_args, training_args = parser.parse_args_into_dataclasses() + + # 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. + send_example_telemetry("run_qa", 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 = training_args.get_process_log_level() + logger.setLevel(log_level) + datasets.utils.logging.set_verbosity(log_level) + transformers.utils.logging.set_verbosity(log_level) + transformers.utils.logging.enable_default_handler() + transformers.utils.logging.enable_explicit_format() + + # Log on each process the small summary: + logger.warning( + f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}" + + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}" + ) + logger.info(f"Training/evaluation parameters {training_args}") + + # Detecting last checkpoint. + last_checkpoint = None + if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir: + last_checkpoint = get_last_checkpoint(training_args.output_dir) + if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0: + raise ValueError( + f"Output directory ({training_args.output_dir}) already exists and is not empty. " + "Use --overwrite_output_dir to overcome." + ) + elif last_checkpoint is not None and training_args.resume_from_checkpoint is None: + logger.info( + f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change " + "the `--output_dir` or add `--overwrite_output_dir` to train from scratch." + ) + + # Set seed before initializing model. + set_seed(training_args.seed) + + # Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below) + # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/ + # (the dataset will be downloaded automatically from the datasets Hub). + # + # For CSV/JSON files, this script will use the column called 'text' or the first column if no column called + # 'text' is found. You can easily tweak this behavior (see below). + # + # In distributed training, the load_dataset function guarantee that only one local process can concurrently + # download the dataset. + cur_dir = os.path.dirname(os.path.abspath(__file__)) + if data_args.dataset_name is not None: + # Downloading and loading a dataset from the hub. + raw_datasets = load_dataset( + os.path.join(cur_dir, "squad_download.py"), + data_args.dataset_config_name, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + else: + data_files = {} + if data_args.train_file is not None: + data_files["train"] = data_args.train_file + extension = data_args.train_file.split(".")[-1] + + if data_args.validation_file is not None: + data_files["validation"] = data_args.validation_file + extension = data_args.validation_file.split(".")[-1] + if data_args.test_file is not None: + data_files["test"] = data_args.test_file + extension = data_args.test_file.split(".")[-1] + raw_datasets = load_dataset( + extension, + data_files=data_files, + field="data", + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + # See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at + # https://huggingface.co/docs/datasets/loading_datasets.html. + + # Load pretrained model and tokenizer + # + # Distributed training: + # The .from_pretrained methods guarantee that only one local process can concurrently + # download model & vocab. + config = AutoConfig.from_pretrained( + model_args.config_name if model_args.config_name else model_args.model_name_or_path, + cache_dir=model_args.cache_dir, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ) + tokenizer = AutoTokenizer.from_pretrained( + model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path, + cache_dir=model_args.cache_dir, + use_fast=True, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ) + model = AutoModelForQuestionAnswering.from_pretrained( + model_args.model_name_or_path, + from_tf=bool(".ckpt" in model_args.model_name_or_path), + config=config, + cache_dir=model_args.cache_dir, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ) + + # Tokenizer check: this script requires a fast tokenizer. + if not isinstance(tokenizer, PreTrainedTokenizerFast): + raise ValueError( + "This example script only works for models that have a fast tokenizer. Checkout the big table of models at" + " https://huggingface.co/transformers/index.html#supported-frameworks to find the model types that meet" + " this requirement" + ) + + # Preprocessing the datasets. + # Preprocessing is slighlty different for training and evaluation. + if training_args.do_train: + column_names = raw_datasets["train"].column_names + elif training_args.do_eval: + column_names = raw_datasets["validation"].column_names + else: + column_names = raw_datasets["test"].column_names + question_column_name = "question" if "question" in column_names else column_names[0] + context_column_name = "context" if "context" in column_names else column_names[1] + answer_column_name = "answers" if "answers" in column_names else column_names[2] + + # Padding side determines if we do (question|context) or (context|question). + pad_on_right = tokenizer.padding_side == "right" + + if data_args.max_seq_length > tokenizer.model_max_length: + logger.warning( + f"The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the" + f"model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}." + ) + max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length) + + # Training preprocessing + def prepare_train_features(examples): + # Some of the questions have lots of whitespace on the left, which is not useful and will make the + # truncation of the context fail (the tokenized question will take a lots of space). So we remove that + # left whitespace + examples[question_column_name] = [q.lstrip() for q in examples[question_column_name]] + + # Tokenize our examples with truncation and maybe padding, but keep the overflows using a stride. This results + # in one example possible giving several features when a context is long, each of those features having a + # context that overlaps a bit the context of the previous feature. + tokenized_examples = tokenizer( + examples[question_column_name if pad_on_right else context_column_name], + examples[context_column_name if pad_on_right else question_column_name], + truncation="only_second" if pad_on_right else "only_first", + max_length=max_seq_length, + stride=data_args.doc_stride, + return_overflowing_tokens=True, + return_offsets_mapping=True, + padding="max_length" if data_args.pad_to_max_length else False, + ) + + # Since one example might give us several features if it has a long context, we need a map from a feature to + # its corresponding example. This key gives us just that. + sample_mapping = tokenized_examples.pop("overflow_to_sample_mapping") + # The offset mappings will give us a map from token to character position in the original context. This will + # help us compute the start_positions and end_positions. + offset_mapping = tokenized_examples.pop("offset_mapping") + + # Let's label those examples! + tokenized_examples["start_positions"] = [] + tokenized_examples["end_positions"] = [] + + for i, offsets in enumerate(offset_mapping): + # We will label impossible answers with the index of the CLS token. + input_ids = tokenized_examples["input_ids"][i] + cls_index = input_ids.index(tokenizer.cls_token_id) + + # Grab the sequence corresponding to that example (to know what is the context and what is the question). + sequence_ids = tokenized_examples.sequence_ids(i) + + # One example can give several spans, this is the index of the example containing this span of text. + sample_index = sample_mapping[i] + answers = examples[answer_column_name][sample_index] + # If no answers are given, set the cls_index as answer. + if len(answers["answer_start"]) == 0: + tokenized_examples["start_positions"].append(cls_index) + tokenized_examples["end_positions"].append(cls_index) + else: + # Start/end character index of the answer in the text. + start_char = answers["answer_start"][0] + end_char = start_char + len(answers["text"][0]) + + # Start token index of the current span in the text. + token_start_index = 0 + while sequence_ids[token_start_index] != (1 if pad_on_right else 0): + token_start_index += 1 + + # End token index of the current span in the text. + token_end_index = len(input_ids) - 1 + while sequence_ids[token_end_index] != (1 if pad_on_right else 0): + token_end_index -= 1 + + # Detect if the answer is out of the span (in which case this feature is labeled with the CLS index). + if not (offsets[token_start_index][0] <= start_char and offsets[token_end_index][1] >= end_char): + tokenized_examples["start_positions"].append(cls_index) + tokenized_examples["end_positions"].append(cls_index) + else: + # Otherwise move the token_start_index and token_end_index to the two ends of the answer. + # Note: we could go after the last offset if the answer is the last word (edge case). + while token_start_index < len(offsets) and offsets[token_start_index][0] <= start_char: + token_start_index += 1 + tokenized_examples["start_positions"].append(token_start_index - 1) + while offsets[token_end_index][1] >= end_char: + token_end_index -= 1 + tokenized_examples["end_positions"].append(token_end_index + 1) + + return tokenized_examples + + if training_args.do_train: + if "train" not in raw_datasets: + raise ValueError("--do_train requires a train dataset") + train_dataset = raw_datasets["train"] + if data_args.max_train_samples is not None: + # We will select sample from whole data if argument is specified + max_train_samples = min(len(train_dataset), data_args.max_train_samples) + train_dataset = train_dataset.select(range(max_train_samples)) + # Create train feature from dataset + with training_args.main_process_first(desc="train dataset map pre-processing"): + train_dataset = train_dataset.map( + prepare_train_features, + batched=True, + num_proc=data_args.preprocessing_num_workers, + remove_columns=column_names, + load_from_cache_file=not data_args.overwrite_cache, + desc="Running tokenizer on train dataset", + ) + if data_args.max_train_samples is not None: + # Number of samples might increase during Feature Creation, We select only specified max samples + max_train_samples = min(len(train_dataset), data_args.max_train_samples) + train_dataset = train_dataset.select(range(max_train_samples)) + + # Validation preprocessing + def prepare_validation_features(examples): + # Some of the questions have lots of whitespace on the left, which is not useful and will make the + # truncation of the context fail (the tokenized question will take a lots of space). So we remove that + # left whitespace + examples[question_column_name] = [q.lstrip() for q in examples[question_column_name]] + + # Tokenize our examples with truncation and maybe padding, but keep the overflows using a stride. This results + # in one example possible giving several features when a context is long, each of those features having a + # context that overlaps a bit the context of the previous feature. + tokenized_examples = tokenizer( + examples[question_column_name if pad_on_right else context_column_name], + examples[context_column_name if pad_on_right else question_column_name], + truncation="only_second" if pad_on_right else "only_first", + max_length=max_seq_length, + stride=data_args.doc_stride, + return_overflowing_tokens=True, + return_offsets_mapping=True, + padding="max_length" if data_args.pad_to_max_length else False, + ) + + # Since one example might give us several features if it has a long context, we need a map from a feature to + # its corresponding example. This key gives us just that. + sample_mapping = tokenized_examples.pop("overflow_to_sample_mapping") + + # For evaluation, we will need to convert our predictions to substrings of the context, so we keep the + # corresponding example_id and we will store the offset mappings. + tokenized_examples["example_id"] = [] + + for i in range(len(tokenized_examples["input_ids"])): + # Grab the sequence corresponding to that example (to know what is the context and what is the question). + sequence_ids = tokenized_examples.sequence_ids(i) + context_index = 1 if pad_on_right else 0 + + # One example can give several spans, this is the index of the example containing this span of text. + sample_index = sample_mapping[i] + tokenized_examples["example_id"].append(examples["id"][sample_index]) + + # Set to None the offset_mapping that are not part of the context so it's easy to determine if a token + # position is part of the context or not. + tokenized_examples["offset_mapping"][i] = [ + (o if sequence_ids[k] == context_index else None) + for k, o in enumerate(tokenized_examples["offset_mapping"][i]) + ] + + return tokenized_examples + + if training_args.do_eval: + if "validation" not in raw_datasets: + raise ValueError("--do_eval requires a validation dataset") + eval_examples = raw_datasets["validation"] + if data_args.max_eval_samples is not None: + # We will select sample from whole data + max_eval_samples = min(len(eval_examples), data_args.max_eval_samples) + eval_examples = eval_examples.select(range(max_eval_samples)) + # Validation Feature Creation + with training_args.main_process_first(desc="validation dataset map pre-processing"): + eval_dataset = eval_examples.map( + prepare_validation_features, + batched=True, + num_proc=data_args.preprocessing_num_workers, + remove_columns=column_names, + load_from_cache_file=not data_args.overwrite_cache, + desc="Running tokenizer on validation dataset", + ) + if data_args.max_eval_samples is not None: + # During Feature creation dataset samples might increase, we will select required samples again + max_eval_samples = min(len(eval_dataset), data_args.max_eval_samples) + eval_dataset = eval_dataset.select(range(max_eval_samples)) + + if training_args.do_predict: + if "test" not in raw_datasets: + raise ValueError("--do_predict requires a test dataset") + predict_examples = raw_datasets["test"] + if data_args.max_predict_samples is not None: + # We will select sample from whole data + predict_examples = predict_examples.select(range(data_args.max_predict_samples)) + # Predict Feature Creation + with training_args.main_process_first(desc="prediction dataset map pre-processing"): + predict_dataset = predict_examples.map( + prepare_validation_features, + batched=True, + num_proc=data_args.preprocessing_num_workers, + remove_columns=column_names, + load_from_cache_file=not data_args.overwrite_cache, + desc="Running tokenizer on prediction dataset", + ) + if data_args.max_predict_samples is not None: + # During Feature creation dataset samples might increase, we will select required samples again + max_predict_samples = min(len(predict_dataset), data_args.max_predict_samples) + predict_dataset = predict_dataset.select(range(max_predict_samples)) + + # Data collator + # We have already padded to max length if the corresponding flag is True, otherwise we need to pad in the data + # collator. + data_collator = ( + default_data_collator + if data_args.pad_to_max_length + else DataCollatorWithPadding(tokenizer, pad_to_multiple_of=8 if training_args.fp16 else None) + ) + + # Post-processing: + def post_processing_function(examples, features, predictions, stage="eval"): + # Post-processing: we match the start logits and end logits to answers in the original context. + predictions = postprocess_qa_predictions( + examples=examples, + features=features, + predictions=predictions, + version_2_with_negative=data_args.version_2_with_negative, + n_best_size=data_args.n_best_size, + max_answer_length=data_args.max_answer_length, + null_score_diff_threshold=data_args.null_score_diff_threshold, + output_dir=training_args.output_dir, + log_level=log_level, + prefix=stage, + ) + # Format the result to the format the metric expects. + if data_args.version_2_with_negative: + formatted_predictions = [ + {"id": k, "prediction_text": v, "no_answer_probability": 0.0} for k, v in predictions.items() + ] + else: + formatted_predictions = [{"id": k, "prediction_text": v} for k, v in predictions.items()] + + references = [{"id": ex["id"], "answers": ex[answer_column_name]} for ex in examples] + return EvalPrediction(predictions=formatted_predictions, label_ids=references) + + metric = evaluate.load("squad_v2" if data_args.version_2_with_negative else "squad") + + def compute_metrics(p: EvalPrediction): + return metric.compute(predictions=p.predictions, references=p.label_ids) + + # Initialize our Trainer + trainer = QuestionAnsweringTrainer( + model=model, + args=training_args, + train_dataset=train_dataset if training_args.do_train else None, + eval_dataset=eval_dataset if training_args.do_eval else None, + eval_examples=eval_examples if training_args.do_eval else None, + tokenizer=tokenizer, + data_collator=data_collator, + post_process_function=post_processing_function, + compute_metrics=compute_metrics, + ) + + # Training + if training_args.do_train: + checkpoint = None + if training_args.resume_from_checkpoint is not None: + checkpoint = training_args.resume_from_checkpoint + elif last_checkpoint is not None: + checkpoint = last_checkpoint + train_result = trainer.train(resume_from_checkpoint=checkpoint) + trainer.save_model() # Saves the tokenizer too for easy upload + + 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() + + # Evaluation + if training_args.do_eval: + logger.info("*** Evaluate ***") + metrics = trainer.evaluate() + + max_eval_samples = data_args.max_eval_samples if data_args.max_eval_samples is not None else len(eval_dataset) + metrics["eval_samples"] = min(max_eval_samples, len(eval_dataset)) + + trainer.log_metrics("eval", metrics) + trainer.save_metrics("eval", metrics) + + # Prediction + if training_args.do_predict: + logger.info("*** Predict ***") + results = trainer.predict(predict_dataset, predict_examples) + metrics = results.metrics + + max_predict_samples = ( + data_args.max_predict_samples if data_args.max_predict_samples is not None else len(predict_dataset) + ) + metrics["predict_samples"] = min(max_predict_samples, len(predict_dataset)) + + trainer.log_metrics("predict", metrics) + trainer.save_metrics("predict", metrics) + + kwargs = {"finetuned_from": model_args.model_name_or_path, "tasks": "question-answering"} + 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) + + +def _mp_fn(index): + # For xla_spawn (TPUs) + main() + + +if __name__ == "__main__": + main() diff --git a/nlp/question_answering/bert/pytorch/squad_download.py b/nlp/question_answering/bert/pytorch/squad_download.py new file mode 100644 index 00000000..76041f92 --- /dev/null +++ b/nlp/question_answering/bert/pytorch/squad_download.py @@ -0,0 +1,153 @@ +# coding=utf-8 +# Copyright 2020 The TensorFlow Datasets Authors and the HuggingFace Datasets Authors. +# +# 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. + +# Lint as: python3 +"""SQUAD: The Stanford Question Answering Dataset.""" + + +import json + +import datasets +from datasets.tasks import QuestionAnsweringExtractive + +logger = datasets.logging.get_logger(__name__) + + +_CITATION = """\ +@article{2016arXiv160605250R, + author = {{Rajpurkar}, Pranav and {Zhang}, Jian and {Lopyrev}, + Konstantin and {Liang}, Percy}, + title = "{SQuAD: 100,000+ Questions for Machine Comprehension of Text}", + journal = {arXiv e-prints}, + year = 2016, + eid = {arXiv:1606.05250}, + pages = {arXiv:1606.05250}, +archivePrefix = {arXiv}, + eprint = {1606.05250}, +} +""" + +_DESCRIPTION = """\ +Stanford Question Answering Dataset (SQuAD) is a reading comprehension \ +dataset, consisting of questions posed by crowdworkers on a set of Wikipedia \ +articles, where the answer to every question is a segment of text, or span, \ +from the corresponding reading passage, or the question might be unanswerable. +""" + +_URL = "https://rajpurkar.github.io/SQuAD-explorer/dataset/" +_URLS = { + "train": _URL + "train-v1.1.json", + "dev": _URL + "dev-v1.1.json", +} + + +class SquadConfig(datasets.BuilderConfig): + """BuilderConfig for SQUAD.""" + + def __init__(self, **kwargs): + """BuilderConfig for SQUAD. + + Args: + **kwargs: keyword arguments forwarded to super. + """ + super(SquadConfig, self).__init__(**kwargs) + + +class Squad(datasets.GeneratorBasedBuilder): + """SQUAD: The Stanford Question Answering Dataset. Version 1.1.""" + + BUILDER_CONFIGS = [ + SquadConfig( + name="plain_text", + version=datasets.Version("1.0.0", ""), + description="Plain text", + ), + ] + + def _info(self): + return datasets.DatasetInfo( + description=_DESCRIPTION, + features=datasets.Features( + { + "id": datasets.Value("string"), + "title": datasets.Value("string"), + "context": datasets.Value("string"), + "question": datasets.Value("string"), + "answers": datasets.features.Sequence( + { + "text": datasets.Value("string"), + "answer_start": datasets.Value("int32"), + } + ), + } + ), + # No default supervised_keys (as we have to pass both question + # and context as input). + supervised_keys=None, + homepage="https://rajpurkar.github.io/SQuAD-explorer/", + citation=_CITATION, + task_templates=[ + QuestionAnsweringExtractive( + question_column="question", + context_column="context", + answers_column="answers", + ) + ], + ) + + def _split_generators(self, dl_manager): + downloaded_files = dl_manager.download_and_extract(_URLS) + + return [ + datasets.SplitGenerator( + name=datasets.Split.TRAIN, + gen_kwargs={"filepath": downloaded_files["train"]}, + ), + datasets.SplitGenerator( + name=datasets.Split.VALIDATION, + gen_kwargs={"filepath": downloaded_files["dev"]}, + ), + ] + + def _generate_examples(self, filepath): + """This function returns the examples in the raw (text) form.""" + logger.info("generating examples from = %s", filepath) + key = 0 + with open(filepath, encoding="utf-8") as f: + squad = json.load(f) + for article in squad["data"]: + title = article.get("title", "") + for paragraph in article["paragraphs"]: + context = paragraph[ + "context" + ] # do not strip leading blank spaces GH-2585 + for qa in paragraph["qas"]: + answer_starts = [ + answer["answer_start"] for answer in qa["answers"] + ] + answers = [answer["text"] for answer in qa["answers"]] + # Features currently used are "context", "question", and "answers". + # Others are extracted here for the ease of future expansions. + yield key, { + "title": title, + "context": context, + "question": qa["question"], + "id": qa["id"], + "answers": { + "answer_start": answer_starts, + "text": answers, + }, + } + key += 1 diff --git a/nlp/question_answering/bert/pytorch/trainer_qa.py b/nlp/question_answering/bert/pytorch/trainer_qa.py new file mode 100644 index 00000000..5535a3fd --- /dev/null +++ b/nlp/question_answering/bert/pytorch/trainer_qa.py @@ -0,0 +1,136 @@ +# coding=utf-8 +# Copyright 2020 The HuggingFace Team All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" +A subclass of `Trainer` specific to Question-Answering tasks +""" +import math +import time + +from transformers import Trainer, is_torch_tpu_available +from transformers.trainer_utils import PredictionOutput, speed_metrics + + +if is_torch_tpu_available(check_device=False): + import torch_xla.core.xla_model as xm + import torch_xla.debug.metrics as met + + +class QuestionAnsweringTrainer(Trainer): + def __init__(self, *args, eval_examples=None, post_process_function=None, **kwargs): + super().__init__(*args, **kwargs) + self.eval_examples = eval_examples + self.post_process_function = post_process_function + + def evaluate(self, eval_dataset=None, eval_examples=None, ignore_keys=None, metric_key_prefix: str = "eval"): + eval_dataset = self.eval_dataset if eval_dataset is None else eval_dataset + eval_dataloader = self.get_eval_dataloader(eval_dataset) + eval_examples = self.eval_examples if eval_examples is None else eval_examples + + # Temporarily disable metric computation, we will do it in the loop here. + compute_metrics = self.compute_metrics + self.compute_metrics = None + eval_loop = self.prediction_loop if self.args.use_legacy_prediction_loop else self.evaluation_loop + start_time = time.time() + try: + output = eval_loop( + eval_dataloader, + description="Evaluation", + # No point gathering the predictions if there are no metrics, otherwise we defer to + # self.args.prediction_loss_only + prediction_loss_only=True if compute_metrics is None else None, + ignore_keys=ignore_keys, + metric_key_prefix=metric_key_prefix, + ) + finally: + self.compute_metrics = compute_metrics + total_batch_size = self.args.eval_batch_size * self.args.world_size + if f"{metric_key_prefix}_jit_compilation_time" in output.metrics: + start_time += output.metrics[f"{metric_key_prefix}_jit_compilation_time"] + output.metrics.update( + speed_metrics( + metric_key_prefix, + start_time, + num_samples=output.num_samples, + num_steps=math.ceil(output.num_samples / total_batch_size), + ) + ) + if self.post_process_function is not None and self.compute_metrics is not None and self.args.should_save: + # Only the main node write the results by default + eval_preds = self.post_process_function(eval_examples, eval_dataset, output.predictions) + metrics = self.compute_metrics(eval_preds) + + # Prefix all keys with metric_key_prefix + '_' + for key in list(metrics.keys()): + if not key.startswith(f"{metric_key_prefix}_"): + metrics[f"{metric_key_prefix}_{key}"] = metrics.pop(key) + metrics.update(output.metrics) + else: + metrics = output.metrics + + if self.args.should_log: + # Only the main node log the results by default + self.log(metrics) + + if self.args.tpu_metrics_debug or self.args.debug: + # tpu-comment: Logging debug metrics for PyTorch/XLA (compile, execute times, ops, etc.) + xm.master_print(met.metrics_report()) + + self.control = self.callback_handler.on_evaluate(self.args, self.state, self.control, metrics) + return metrics + + def predict(self, predict_dataset, predict_examples, ignore_keys=None, metric_key_prefix: str = "test"): + predict_dataloader = self.get_test_dataloader(predict_dataset) + + # Temporarily disable metric computation, we will do it in the loop here. + compute_metrics = self.compute_metrics + self.compute_metrics = None + eval_loop = self.prediction_loop if self.args.use_legacy_prediction_loop else self.evaluation_loop + start_time = time.time() + try: + output = eval_loop( + predict_dataloader, + description="Prediction", + # No point gathering the predictions if there are no metrics, otherwise we defer to + # self.args.prediction_loss_only + prediction_loss_only=True if compute_metrics is None else None, + ignore_keys=ignore_keys, + metric_key_prefix=metric_key_prefix, + ) + finally: + self.compute_metrics = compute_metrics + total_batch_size = self.args.eval_batch_size * self.args.world_size + if f"{metric_key_prefix}_jit_compilation_time" in output.metrics: + start_time += output.metrics[f"{metric_key_prefix}_jit_compilation_time"] + output.metrics.update( + speed_metrics( + metric_key_prefix, + start_time, + num_samples=output.num_samples, + num_steps=math.ceil(output.num_samples / total_batch_size), + ) + ) + + if self.post_process_function is None or self.compute_metrics is None: + return output + + predictions = self.post_process_function(predict_examples, predict_dataset, output.predictions, "predict") + metrics = self.compute_metrics(predictions) + + # Prefix all keys with metric_key_prefix + '_' + for key in list(metrics.keys()): + if not key.startswith(f"{metric_key_prefix}_"): + metrics[f"{metric_key_prefix}_{key}"] = metrics.pop(key) + metrics.update(output.metrics) + return PredictionOutput(predictions=predictions.predictions, label_ids=predictions.label_ids, metrics=metrics) \ No newline at end of file diff --git a/nlp/question_answering/bert/pytorch/utils_qa.py b/nlp/question_answering/bert/pytorch/utils_qa.py new file mode 100644 index 00000000..345e0dbd --- /dev/null +++ b/nlp/question_answering/bert/pytorch/utils_qa.py @@ -0,0 +1,443 @@ +# coding=utf-8 +# Copyright 2020 The HuggingFace Team All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" +Post-processing utilities for question answering. +""" +import collections +import json +import logging +import os +from typing import Optional, Tuple + +import numpy as np +from tqdm.auto import tqdm + + +logger = logging.getLogger(__name__) + + +def postprocess_qa_predictions( + examples, + features, + predictions: Tuple[np.ndarray, np.ndarray], + version_2_with_negative: bool = False, + n_best_size: int = 20, + max_answer_length: int = 30, + null_score_diff_threshold: float = 0.0, + output_dir: Optional[str] = None, + prefix: Optional[str] = None, + log_level: Optional[int] = logging.WARNING, +): + """ + Post-processes the predictions of a question-answering model to convert them to answers that are substrings of the + original contexts. This is the base postprocessing functions for models that only return start and end logits. + + Args: + examples: The non-preprocessed dataset (see the main script for more information). + features: The processed dataset (see the main script for more information). + predictions (:obj:`Tuple[np.ndarray, np.ndarray]`): + The predictions of the model: two arrays containing the start logits and the end logits respectively. Its + first dimension must match the number of elements of :obj:`features`. + version_2_with_negative (:obj:`bool`, `optional`, defaults to :obj:`False`): + Whether or not the underlying dataset contains examples with no answers. + n_best_size (:obj:`int`, `optional`, defaults to 20): + The total number of n-best predictions to generate when looking for an answer. + max_answer_length (:obj:`int`, `optional`, defaults to 30): + 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. + null_score_diff_threshold (:obj:`float`, `optional`, defaults to 0): + The threshold used to select the null answer: if the best answer has a score that is less than the score of + the null answer minus this threshold, the null answer is selected for this example (note that the score of + the null answer for an example giving several features is the minimum of the scores for the null answer on + each feature: all features must be aligned on the fact they `want` to predict a null answer). + + Only useful when :obj:`version_2_with_negative` is :obj:`True`. + output_dir (:obj:`str`, `optional`): + If provided, the dictionaries of predictions, n_best predictions (with their scores and logits) and, if + :obj:`version_2_with_negative=True`, the dictionary of the scores differences between best and null + answers, are saved in `output_dir`. + prefix (:obj:`str`, `optional`): + If provided, the dictionaries mentioned above are saved with `prefix` added to their names. + log_level (:obj:`int`, `optional`, defaults to ``logging.WARNING``): + ``logging`` log level (e.g., ``logging.WARNING``) + """ + if len(predictions) != 2: + raise ValueError("`predictions` should be a tuple with two elements (start_logits, end_logits).") + all_start_logits, all_end_logits = predictions + + if len(predictions[0]) != len(features): + raise ValueError(f"Got {len(predictions[0])} predictions and {len(features)} features.") + + # Build a map example to its corresponding features. + example_id_to_index = {k: i for i, k in enumerate(examples["id"])} + features_per_example = collections.defaultdict(list) + for i, feature in enumerate(features): + features_per_example[example_id_to_index[feature["example_id"]]].append(i) + + # The dictionaries we have to fill. + all_predictions = collections.OrderedDict() + all_nbest_json = collections.OrderedDict() + if version_2_with_negative: + scores_diff_json = collections.OrderedDict() + + # Logging. + logger.setLevel(log_level) + logger.info(f"Post-processing {len(examples)} example predictions split into {len(features)} features.") + + # Let's loop over all the examples! + for example_index, example in enumerate(tqdm(examples)): + # Those are the indices of the features associated to the current example. + feature_indices = features_per_example[example_index] + + min_null_prediction = None + prelim_predictions = [] + + # Looping through all the features associated to the current example. + for feature_index in feature_indices: + # We grab the predictions of the model for this feature. + start_logits = all_start_logits[feature_index] + end_logits = all_end_logits[feature_index] + # This is what will allow us to map some the positions in our logits to span of texts in the original + # context. + offset_mapping = features[feature_index]["offset_mapping"] + # Optional `token_is_max_context`, if provided we will remove answers that do not have the maximum context + # available in the current feature. + token_is_max_context = features[feature_index].get("token_is_max_context", None) + + # Update minimum null prediction. + feature_null_score = start_logits[0] + end_logits[0] + if min_null_prediction is None or min_null_prediction["score"] > feature_null_score: + min_null_prediction = { + "offsets": (0, 0), + "score": feature_null_score, + "start_logit": start_logits[0], + "end_logit": end_logits[0], + } + + # Go through all possibilities for the `n_best_size` greater start and end logits. + start_indexes = np.argsort(start_logits)[-1 : -n_best_size - 1 : -1].tolist() + end_indexes = np.argsort(end_logits)[-1 : -n_best_size - 1 : -1].tolist() + for start_index in start_indexes: + for end_index in end_indexes: + # Don't consider out-of-scope answers, either because the indices are out of bounds or correspond + # to part of the input_ids that are not in the context. + if ( + start_index >= len(offset_mapping) + or end_index >= len(offset_mapping) + or offset_mapping[start_index] is None + or len(offset_mapping[start_index]) < 2 + or offset_mapping[end_index] is None + or len(offset_mapping[end_index]) < 2 + ): + continue + # Don't consider answers with a length that is either < 0 or > max_answer_length. + if end_index < start_index or end_index - start_index + 1 > max_answer_length: + continue + # Don't consider answer that don't have the maximum context available (if such information is + # provided). + if token_is_max_context is not None and not token_is_max_context.get(str(start_index), False): + continue + + prelim_predictions.append( + { + "offsets": (offset_mapping[start_index][0], offset_mapping[end_index][1]), + "score": start_logits[start_index] + end_logits[end_index], + "start_logit": start_logits[start_index], + "end_logit": end_logits[end_index], + } + ) + if version_2_with_negative and min_null_prediction is not None: + # Add the minimum null prediction + prelim_predictions.append(min_null_prediction) + null_score = min_null_prediction["score"] + + # Only keep the best `n_best_size` predictions. + predictions = sorted(prelim_predictions, key=lambda x: x["score"], reverse=True)[:n_best_size] + + # Add back the minimum null prediction if it was removed because of its low score. + if ( + version_2_with_negative + and min_null_prediction is not None + and not any(p["offsets"] == (0, 0) for p in predictions) + ): + predictions.append(min_null_prediction) + + # Use the offsets to gather the answer text in the original context. + context = example["context"] + for pred in predictions: + offsets = pred.pop("offsets") + pred["text"] = context[offsets[0] : offsets[1]] + + # In the very rare edge case we have not a single non-null prediction, we create a fake prediction to avoid + # failure. + if len(predictions) == 0 or (len(predictions) == 1 and predictions[0]["text"] == ""): + predictions.insert(0, {"text": "empty", "start_logit": 0.0, "end_logit": 0.0, "score": 0.0}) + + # Compute the softmax of all scores (we do it with numpy to stay independent from torch/tf in this file, using + # the LogSumExp trick). + scores = np.array([pred.pop("score") for pred in predictions]) + exp_scores = np.exp(scores - np.max(scores)) + probs = exp_scores / exp_scores.sum() + + # Include the probabilities in our predictions. + for prob, pred in zip(probs, predictions): + pred["probability"] = prob + + # Pick the best prediction. If the null answer is not possible, this is easy. + if not version_2_with_negative: + all_predictions[example["id"]] = predictions[0]["text"] + else: + # Otherwise we first need to find the best non-empty prediction. + i = 0 + while predictions[i]["text"] == "": + i += 1 + best_non_null_pred = predictions[i] + + # Then we compare to the null prediction using the threshold. + score_diff = null_score - best_non_null_pred["start_logit"] - best_non_null_pred["end_logit"] + scores_diff_json[example["id"]] = float(score_diff) # To be JSON-serializable. + if score_diff > null_score_diff_threshold: + all_predictions[example["id"]] = "" + else: + all_predictions[example["id"]] = best_non_null_pred["text"] + + # Make `predictions` JSON-serializable by casting np.float back to float. + all_nbest_json[example["id"]] = [ + {k: (float(v) if isinstance(v, (np.float16, np.float32, np.float64)) else v) for k, v in pred.items()} + for pred in predictions + ] + + # If we have an output_dir, let's save all those dicts. + if output_dir is not None: + if not os.path.isdir(output_dir): + raise EnvironmentError(f"{output_dir} is not a directory.") + + prediction_file = os.path.join( + output_dir, "predictions.json" if prefix is None else f"{prefix}_predictions.json" + ) + nbest_file = os.path.join( + output_dir, "nbest_predictions.json" if prefix is None else f"{prefix}_nbest_predictions.json" + ) + if version_2_with_negative: + null_odds_file = os.path.join( + output_dir, "null_odds.json" if prefix is None else f"{prefix}_null_odds.json" + ) + + logger.info(f"Saving predictions to {prediction_file}.") + with open(prediction_file, "w") as writer: + writer.write(json.dumps(all_predictions, indent=4) + "\n") + logger.info(f"Saving nbest_preds to {nbest_file}.") + with open(nbest_file, "w") as writer: + writer.write(json.dumps(all_nbest_json, indent=4) + "\n") + if version_2_with_negative: + logger.info(f"Saving null_odds to {null_odds_file}.") + with open(null_odds_file, "w") as writer: + writer.write(json.dumps(scores_diff_json, indent=4) + "\n") + + return all_predictions + + +def postprocess_qa_predictions_with_beam_search( + examples, + features, + predictions: Tuple[np.ndarray, np.ndarray], + version_2_with_negative: bool = False, + n_best_size: int = 20, + max_answer_length: int = 30, + start_n_top: int = 5, + end_n_top: int = 5, + output_dir: Optional[str] = None, + prefix: Optional[str] = None, + log_level: Optional[int] = logging.WARNING, +): + """ + Post-processes the predictions of a question-answering model with beam search to convert them to answers that are substrings of the + original contexts. This is the postprocessing functions for models that return start and end logits, indices, as well as + cls token predictions. + + Args: + examples: The non-preprocessed dataset (see the main script for more information). + features: The processed dataset (see the main script for more information). + predictions (:obj:`Tuple[np.ndarray, np.ndarray]`): + The predictions of the model: two arrays containing the start logits and the end logits respectively. Its + first dimension must match the number of elements of :obj:`features`. + version_2_with_negative (:obj:`bool`, `optional`, defaults to :obj:`False`): + Whether or not the underlying dataset contains examples with no answers. + n_best_size (:obj:`int`, `optional`, defaults to 20): + The total number of n-best predictions to generate when looking for an answer. + max_answer_length (:obj:`int`, `optional`, defaults to 30): + 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. + start_n_top (:obj:`int`, `optional`, defaults to 5): + The number of top start logits too keep when searching for the :obj:`n_best_size` predictions. + end_n_top (:obj:`int`, `optional`, defaults to 5): + The number of top end logits too keep when searching for the :obj:`n_best_size` predictions. + output_dir (:obj:`str`, `optional`): + If provided, the dictionaries of predictions, n_best predictions (with their scores and logits) and, if + :obj:`version_2_with_negative=True`, the dictionary of the scores differences between best and null + answers, are saved in `output_dir`. + prefix (:obj:`str`, `optional`): + If provided, the dictionaries mentioned above are saved with `prefix` added to their names. + log_level (:obj:`int`, `optional`, defaults to ``logging.WARNING``): + ``logging`` log level (e.g., ``logging.WARNING``) + """ + if len(predictions) != 5: + raise ValueError("`predictions` should be a tuple with five elements.") + start_top_log_probs, start_top_index, end_top_log_probs, end_top_index, cls_logits = predictions + + if len(predictions[0]) != len(features): + raise ValueError(f"Got {len(predictions[0])} predictions and {len(features)} features.") + + # Build a map example to its corresponding features. + example_id_to_index = {k: i for i, k in enumerate(examples["id"])} + features_per_example = collections.defaultdict(list) + for i, feature in enumerate(features): + features_per_example[example_id_to_index[feature["example_id"]]].append(i) + + # The dictionaries we have to fill. + all_predictions = collections.OrderedDict() + all_nbest_json = collections.OrderedDict() + scores_diff_json = collections.OrderedDict() if version_2_with_negative else None + + # Logging. + logger.setLevel(log_level) + logger.info(f"Post-processing {len(examples)} example predictions split into {len(features)} features.") + + # Let's loop over all the examples! + for example_index, example in enumerate(tqdm(examples)): + # Those are the indices of the features associated to the current example. + feature_indices = features_per_example[example_index] + + min_null_score = None + prelim_predictions = [] + + # Looping through all the features associated to the current example. + for feature_index in feature_indices: + # We grab the predictions of the model for this feature. + start_log_prob = start_top_log_probs[feature_index] + start_indexes = start_top_index[feature_index] + end_log_prob = end_top_log_probs[feature_index] + end_indexes = end_top_index[feature_index] + feature_null_score = cls_logits[feature_index] + # This is what will allow us to map some the positions in our logits to span of texts in the original + # context. + offset_mapping = features[feature_index]["offset_mapping"] + # Optional `token_is_max_context`, if provided we will remove answers that do not have the maximum context + # available in the current feature. + token_is_max_context = features[feature_index].get("token_is_max_context", None) + + # Update minimum null prediction + if min_null_score is None or feature_null_score < min_null_score: + min_null_score = feature_null_score + + # Go through all possibilities for the `n_start_top`/`n_end_top` greater start and end logits. + for i in range(start_n_top): + for j in range(end_n_top): + start_index = int(start_indexes[i]) + j_index = i * end_n_top + j + end_index = int(end_indexes[j_index]) + # Don't consider out-of-scope answers (last part of the test should be unnecessary because of the + # p_mask but let's not take any risk) + if ( + start_index >= len(offset_mapping) + or end_index >= len(offset_mapping) + or offset_mapping[start_index] is None + or len(offset_mapping[start_index]) < 2 + or offset_mapping[end_index] is None + or len(offset_mapping[end_index]) < 2 + ): + continue + + # Don't consider answers with a length negative or > max_answer_length. + if end_index < start_index or end_index - start_index + 1 > max_answer_length: + continue + # Don't consider answer that don't have the maximum context available (if such information is + # provided). + if token_is_max_context is not None and not token_is_max_context.get(str(start_index), False): + continue + prelim_predictions.append( + { + "offsets": (offset_mapping[start_index][0], offset_mapping[end_index][1]), + "score": start_log_prob[i] + end_log_prob[j_index], + "start_log_prob": start_log_prob[i], + "end_log_prob": end_log_prob[j_index], + } + ) + + # Only keep the best `n_best_size` predictions. + predictions = sorted(prelim_predictions, key=lambda x: x["score"], reverse=True)[:n_best_size] + + # Use the offsets to gather the answer text in the original context. + context = example["context"] + for pred in predictions: + offsets = pred.pop("offsets") + pred["text"] = context[offsets[0] : offsets[1]] + + # In the very rare edge case we have not a single non-null prediction, we create a fake prediction to avoid + # failure. + if len(predictions) == 0: + # Without predictions min_null_score is going to be None and None will cause an exception later + min_null_score = -2e-6 + predictions.insert(0, {"text": "", "start_logit": -1e-6, "end_logit": -1e-6, "score": min_null_score}) + + # Compute the softmax of all scores (we do it with numpy to stay independent from torch/tf in this file, using + # the LogSumExp trick). + scores = np.array([pred.pop("score") for pred in predictions]) + exp_scores = np.exp(scores - np.max(scores)) + probs = exp_scores / exp_scores.sum() + + # Include the probabilities in our predictions. + for prob, pred in zip(probs, predictions): + pred["probability"] = prob + + # Pick the best prediction and set the probability for the null answer. + all_predictions[example["id"]] = predictions[0]["text"] + if version_2_with_negative: + scores_diff_json[example["id"]] = float(min_null_score) + + # Make `predictions` JSON-serializable by casting np.float back to float. + all_nbest_json[example["id"]] = [ + {k: (float(v) if isinstance(v, (np.float16, np.float32, np.float64)) else v) for k, v in pred.items()} + for pred in predictions + ] + + # If we have an output_dir, let's save all those dicts. + if output_dir is not None: + if not os.path.isdir(output_dir): + raise EnvironmentError(f"{output_dir} is not a directory.") + + prediction_file = os.path.join( + output_dir, "predictions.json" if prefix is None else f"{prefix}_predictions.json" + ) + nbest_file = os.path.join( + output_dir, "nbest_predictions.json" if prefix is None else f"{prefix}_nbest_predictions.json" + ) + if version_2_with_negative: + null_odds_file = os.path.join( + output_dir, "null_odds.json" if prefix is None else f"{prefix}_null_odds.json" + ) + + logger.info(f"Saving predictions to {prediction_file}.") + with open(prediction_file, "w") as writer: + writer.write(json.dumps(all_predictions, indent=4) + "\n") + logger.info(f"Saving nbest_preds to {nbest_file}.") + with open(nbest_file, "w") as writer: + writer.write(json.dumps(all_nbest_json, indent=4) + "\n") + if version_2_with_negative: + logger.info(f"Saving null_odds to {null_odds_file}.") + with open(null_odds_file, "w") as writer: + writer.write(json.dumps(scores_diff_json, indent=4) + "\n") + + return all_predictions, scores_diff_json \ No newline at end of file -- Gitee From 4c29e9f1b14bf100b74f8c0a34c333faa5a701d4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E7=89=9B=E5=86=A0=E5=8D=9A?= Date: Fri, 24 Feb 2023 16:55:45 +0800 Subject: [PATCH 2/4] adjust gitignore --- .gitignore | 2 ++ 1 file changed, 2 insertions(+) diff --git a/.gitignore b/.gitignore index 5c452d07..c9406253 100644 --- a/.gitignore +++ b/.gitignore @@ -13,3 +13,5 @@ dist/ # Pytorch *.pth + +*.idea -- Gitee From 7f882326b17ae278752fd8abefcee6f95242d3ce Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E7=89=9B=E5=86=A0=E5=8D=9A?= Date: Fri, 24 Feb 2023 22:59:37 +0800 Subject: [PATCH 3/4] add bert base text_classification source code --- nlp/text_classification/README.md | 36 ++ nlp/text_classification/requirements.txt | 10 + nlp/text_classification/run_glue.py | 622 +++++++++++++++++++++++ nlp/text_classification/train.sh | 12 + nlp/text_classification/train_dist.sh | 12 + 5 files changed, 692 insertions(+) create mode 100644 nlp/text_classification/requirements.txt create mode 100644 nlp/text_classification/run_glue.py create mode 100644 nlp/text_classification/train.sh create mode 100644 nlp/text_classification/train_dist.sh diff --git a/nlp/text_classification/README.md b/nlp/text_classification/README.md index a4683851..c82b996d 100644 --- a/nlp/text_classification/README.md +++ b/nlp/text_classification/README.md @@ -1 +1,37 @@ # Text Classification + +# Bert-base WNLI + +## Model description + +Bert-base WNLI task Fine-tuning + +## Step 1: Installing packages + +``` shell +cd /nlp/text_classification/bert/pytorch +pip3 install -r requirements.txt +``` + +## Step 2: Training + +### On single GPU + +``` shell +bash train.sh +``` + +### Multiple GPUs on one machine + +```shell +bash train_dist.sh +``` +## Results on BI-V100 + +| GPUs | Samples/s | Loss | +|------|-----------|------| +| 1x1 | 144.5 | 0.74 | +| 1x8 | 322.74 | 0.71 | + +## Reference +https://github.com/huggingface/ diff --git a/nlp/text_classification/requirements.txt b/nlp/text_classification/requirements.txt new file mode 100644 index 00000000..dbfdb439 --- /dev/null +++ b/nlp/text_classification/requirements.txt @@ -0,0 +1,10 @@ +accelerate >= 0.12.0 +datasets >= 1.8.0 +sentencepiece != 0.1.92 +scipy +scikit-learn +protobuf +# torch >= 1.3 +evaluate +numpy == 1.23.4 +git+https://github.com/huggingface/transformers \ No newline at end of file diff --git a/nlp/text_classification/run_glue.py b/nlp/text_classification/run_glue.py new file mode 100644 index 00000000..e8ede80e --- /dev/null +++ b/nlp/text_classification/run_glue.py @@ -0,0 +1,622 @@ +#!/usr/bin/env python +# coding=utf-8 +# Copyright 2020 The HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" Finetuning the library models for sequence classification on GLUE.""" +# You can also adapt this script on your own text classification task. Pointers for this are left as comments. + +import logging +import os +import random +import sys +from dataclasses import dataclass, field +from typing import Optional + +import datasets +import evaluate +import numpy as np +from datasets import load_dataset + +import transformers +from transformers import ( + AutoConfig, + AutoModelForSequenceClassification, + AutoTokenizer, + DataCollatorWithPadding, + EvalPrediction, + HfArgumentParser, + PretrainedConfig, + Trainer, + TrainingArguments, + default_data_collator, + set_seed, +) +from transformers.trainer_utils import get_last_checkpoint +from transformers.utils import check_min_version, send_example_telemetry +from transformers.utils.versions import require_version + + +# Will error if the minimal version of Transformers is not installed. Remove at your own risks. +check_min_version("4.27.0.dev0") + +require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/text-classification/requirements.txt") + +task_to_keys = { + "cola": ("sentence", None), + "mnli": ("premise", "hypothesis"), + "mrpc": ("sentence1", "sentence2"), + "qnli": ("question", "sentence"), + "qqp": ("question1", "question2"), + "rte": ("sentence1", "sentence2"), + "sst2": ("sentence", None), + "stsb": ("sentence1", "sentence2"), + "wnli": ("sentence1", "sentence2"), +} + +logger = logging.getLogger(__name__) + + +@dataclass +class DataTrainingArguments: + """ + Arguments pertaining to what data we are going to input our model for training and eval. + + Using `HfArgumentParser` we can turn this class + into argparse arguments to be able to specify them on + the command line. + """ + + task_name: Optional[str] = field( + default=None, + metadata={"help": "The name of the task to train on: " + ", ".join(task_to_keys.keys())}, + ) + dataset_name: Optional[str] = field( + default=None, metadata={"help": "The name of the dataset to use (via the datasets library)."} + ) + dataset_config_name: Optional[str] = field( + default=None, metadata={"help": "The configuration name of the dataset to use (via the datasets library)."} + ) + max_seq_length: int = field( + default=128, + metadata={ + "help": ( + "The maximum total input sequence length after tokenization. Sequences longer " + "than this will be truncated, sequences shorter will be padded." + ) + }, + ) + overwrite_cache: bool = field( + default=False, metadata={"help": "Overwrite the cached preprocessed datasets or not."} + ) + pad_to_max_length: bool = field( + default=True, + metadata={ + "help": ( + "Whether to pad all samples to `max_seq_length`. " + "If False, will pad the samples dynamically when batching to the maximum length in the batch." + ) + }, + ) + max_train_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of training examples to this " + "value if set." + ) + }, + ) + max_eval_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of evaluation examples to this " + "value if set." + ) + }, + ) + max_predict_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of prediction examples to this " + "value if set." + ) + }, + ) + train_file: Optional[str] = field( + default=None, metadata={"help": "A csv or a json file containing the training data."} + ) + validation_file: Optional[str] = field( + default=None, metadata={"help": "A csv or a json file containing the validation data."} + ) + test_file: Optional[str] = field(default=None, metadata={"help": "A csv or a json file containing the test data."}) + + def __post_init__(self): + if self.task_name is not None: + self.task_name = self.task_name.lower() + if self.task_name not in task_to_keys.keys(): + raise ValueError("Unknown task, you should pick one in " + ",".join(task_to_keys.keys())) + elif self.dataset_name is not None: + pass + elif self.train_file is None or self.validation_file is None: + raise ValueError("Need either a GLUE task, a training/validation file or a dataset name.") + else: + train_extension = self.train_file.split(".")[-1] + assert train_extension in ["csv", "json"], "`train_file` should be a csv or a json file." + validation_extension = self.validation_file.split(".")[-1] + assert ( + validation_extension == train_extension + ), "`validation_file` should have the same extension (csv or json) as `train_file`." + + +@dataclass +class ModelArguments: + """ + Arguments pertaining to which model/config/tokenizer we are going to fine-tune from. + """ + + model_name_or_path: str = field( + metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"} + ) + config_name: Optional[str] = field( + default=None, metadata={"help": "Pretrained config name or path if not the same as model_name"} + ) + tokenizer_name: Optional[str] = field( + default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"} + ) + cache_dir: Optional[str] = field( + default=None, + metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}, + ) + use_fast_tokenizer: bool = field( + default=True, + metadata={"help": "Whether to use one of the fast tokenizer (backed by the tokenizers library) or not."}, + ) + model_revision: str = field( + default="main", + metadata={"help": "The specific model version to use (can be a branch name, tag name or commit id)."}, + ) + use_auth_token: bool = field( + default=False, + metadata={ + "help": ( + "Will use the token generated when running `huggingface-cli login` (necessary to use this script " + "with private models)." + ) + }, + ) + ignore_mismatched_sizes: bool = field( + default=False, + metadata={"help": "Will enable to load a pretrained model whose head dimensions are different."}, + ) + + +def main(): + # See all possible arguments in src/transformers/training_args.py + # or by passing the --help flag to this script. + # We now keep distinct sets of args, for a cleaner separation of concerns. + + parser = HfArgumentParser((ModelArguments, DataTrainingArguments, TrainingArguments)) + if len(sys.argv) == 2 and sys.argv[1].endswith(".json"): + # If we pass only one argument to the script and it's the path to a json file, + # let's parse it to get our arguments. + model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1])) + else: + model_args, data_args, training_args = parser.parse_args_into_dataclasses() + + # 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. + send_example_telemetry("run_glue", 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 = training_args.get_process_log_level() + logger.setLevel(log_level) + datasets.utils.logging.set_verbosity(log_level) + transformers.utils.logging.set_verbosity(log_level) + transformers.utils.logging.enable_default_handler() + transformers.utils.logging.enable_explicit_format() + + # Log on each process the small summary: + logger.warning( + f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}" + + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}" + ) + logger.info(f"Training/evaluation parameters {training_args}") + + # Detecting last checkpoint. + last_checkpoint = None + if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir: + last_checkpoint = get_last_checkpoint(training_args.output_dir) + if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0: + raise ValueError( + f"Output directory ({training_args.output_dir}) already exists and is not empty. " + "Use --overwrite_output_dir to overcome." + ) + elif last_checkpoint is not None and training_args.resume_from_checkpoint is None: + logger.info( + f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change " + "the `--output_dir` or add `--overwrite_output_dir` to train from scratch." + ) + + # Set seed before initializing model. + set_seed(training_args.seed) + + # Get the datasets: you can either provide your own CSV/JSON training and evaluation files (see below) + # or specify a GLUE benchmark task (the dataset will be downloaded automatically from the datasets Hub). + # + # For CSV/JSON files, this script will use as labels the column called 'label' and as pair of sentences the + # sentences in columns called 'sentence1' and 'sentence2' if such column exists or the first two columns not named + # label if at least two columns are provided. + # + # If the CSVs/JSONs contain only one non-label column, the script does single sentence classification on this + # single column. You can easily tweak this behavior (see below) + # + # In distributed training, the load_dataset function guarantee that only one local process can concurrently + # download the dataset. + if data_args.task_name is not None: + # Downloading and loading a dataset from the hub. + raw_datasets = load_dataset( + "glue", + data_args.task_name, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + elif data_args.dataset_name is not None: + # Downloading and loading a dataset from the hub. + raw_datasets = load_dataset( + data_args.dataset_name, + data_args.dataset_config_name, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + else: + # Loading a dataset from your local files. + # CSV/JSON training and evaluation files are needed. + data_files = {"train": data_args.train_file, "validation": data_args.validation_file} + + # Get the test dataset: you can provide your own CSV/JSON test file (see below) + # when you use `do_predict` without specifying a GLUE benchmark task. + if training_args.do_predict: + if data_args.test_file is not None: + train_extension = data_args.train_file.split(".")[-1] + test_extension = data_args.test_file.split(".")[-1] + assert ( + test_extension == train_extension + ), "`test_file` should have the same extension (csv or json) as `train_file`." + data_files["test"] = data_args.test_file + else: + raise ValueError("Need either a GLUE task or a test file for `do_predict`.") + + for key in data_files.keys(): + logger.info(f"load a local file for {key}: {data_files[key]}") + + if data_args.train_file.endswith(".csv"): + # Loading a dataset from local csv files + raw_datasets = load_dataset( + "csv", + data_files=data_files, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + else: + # Loading a dataset from local json files + raw_datasets = load_dataset( + "json", + data_files=data_files, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + # See more about loading any type of standard or custom dataset at + # https://huggingface.co/docs/datasets/loading_datasets.html. + + # Labels + if data_args.task_name is not None: + is_regression = data_args.task_name == "stsb" + if not is_regression: + label_list = raw_datasets["train"].features["label"].names + num_labels = len(label_list) + else: + num_labels = 1 + else: + # Trying to have good defaults here, don't hesitate to tweak to your needs. + is_regression = raw_datasets["train"].features["label"].dtype in ["float32", "float64"] + if is_regression: + num_labels = 1 + else: + # A useful fast method: + # https://huggingface.co/docs/datasets/package_reference/main_classes.html#datasets.Dataset.unique + label_list = raw_datasets["train"].unique("label") + label_list.sort() # Let's sort it for determinism + num_labels = len(label_list) + + # Load pretrained model and tokenizer + # + # In distributed training, the .from_pretrained methods guarantee that only one local process can concurrently + # download model & vocab. + config = AutoConfig.from_pretrained( + model_args.config_name if model_args.config_name else model_args.model_name_or_path, + num_labels=num_labels, + finetuning_task=data_args.task_name, + cache_dir=model_args.cache_dir, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ) + tokenizer = AutoTokenizer.from_pretrained( + model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path, + cache_dir=model_args.cache_dir, + use_fast=model_args.use_fast_tokenizer, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ) + model = AutoModelForSequenceClassification.from_pretrained( + model_args.model_name_or_path, + from_tf=bool(".ckpt" in model_args.model_name_or_path), + config=config, + cache_dir=model_args.cache_dir, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ignore_mismatched_sizes=model_args.ignore_mismatched_sizes, + ) + + # Preprocessing the raw_datasets + if data_args.task_name is not None: + sentence1_key, sentence2_key = task_to_keys[data_args.task_name] + else: + # Again, we try to have some nice defaults but don't hesitate to tweak to your use case. + non_label_column_names = [name for name in raw_datasets["train"].column_names if name != "label"] + if "sentence1" in non_label_column_names and "sentence2" in non_label_column_names: + sentence1_key, sentence2_key = "sentence1", "sentence2" + else: + if len(non_label_column_names) >= 2: + sentence1_key, sentence2_key = non_label_column_names[:2] + else: + sentence1_key, sentence2_key = non_label_column_names[0], None + + # Padding strategy + if data_args.pad_to_max_length: + padding = "max_length" + else: + # We will pad later, dynamically at batch creation, to the max sequence length in each batch + padding = False + + # Some models have set the order of the labels to use, so let's make sure we do use it. + label_to_id = None + if ( + model.config.label2id != PretrainedConfig(num_labels=num_labels).label2id + and data_args.task_name is not None + and not is_regression + ): + # Some have all caps in their config, some don't. + label_name_to_id = {k.lower(): v for k, v in model.config.label2id.items()} + if list(sorted(label_name_to_id.keys())) == list(sorted(label_list)): + label_to_id = {i: int(label_name_to_id[label_list[i]]) for i in range(num_labels)} + else: + logger.warning( + "Your model seems to have been trained with labels, but they don't match the dataset: ", + f"model labels: {list(sorted(label_name_to_id.keys()))}, dataset labels: {list(sorted(label_list))}." + "\nIgnoring the model labels as a result.", + ) + elif data_args.task_name is None and not is_regression: + label_to_id = {v: i for i, v in enumerate(label_list)} + + if label_to_id is not None: + model.config.label2id = label_to_id + model.config.id2label = {id: label for label, id in config.label2id.items()} + elif data_args.task_name is not None and not is_regression: + model.config.label2id = {l: i for i, l in enumerate(label_list)} + model.config.id2label = {id: label for label, id in config.label2id.items()} + + if data_args.max_seq_length > tokenizer.model_max_length: + logger.warning( + f"The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the" + f"model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}." + ) + max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length) + + def preprocess_function(examples): + # Tokenize the texts + args = ( + (examples[sentence1_key],) if sentence2_key is None else (examples[sentence1_key], examples[sentence2_key]) + ) + result = tokenizer(*args, padding=padding, max_length=max_seq_length, truncation=True) + + # Map labels to IDs (not necessary for GLUE tasks) + if label_to_id is not None and "label" in examples: + result["label"] = [(label_to_id[l] if l != -1 else -1) for l in examples["label"]] + return result + + with training_args.main_process_first(desc="dataset map pre-processing"): + raw_datasets = raw_datasets.map( + preprocess_function, + batched=True, + load_from_cache_file=not data_args.overwrite_cache, + desc="Running tokenizer on dataset", + ) + if training_args.do_train: + if "train" not in raw_datasets: + raise ValueError("--do_train requires a train dataset") + train_dataset = raw_datasets["train"] + if data_args.max_train_samples is not None: + max_train_samples = min(len(train_dataset), data_args.max_train_samples) + train_dataset = train_dataset.select(range(max_train_samples)) + + if training_args.do_eval: + if "validation" not in raw_datasets and "validation_matched" not in raw_datasets: + raise ValueError("--do_eval requires a validation dataset") + eval_dataset = raw_datasets["validation_matched" if data_args.task_name == "mnli" else "validation"] + if data_args.max_eval_samples is not None: + max_eval_samples = min(len(eval_dataset), data_args.max_eval_samples) + eval_dataset = eval_dataset.select(range(max_eval_samples)) + + if training_args.do_predict or data_args.task_name is not None or data_args.test_file is not None: + if "test" not in raw_datasets and "test_matched" not in raw_datasets: + raise ValueError("--do_predict requires a test dataset") + predict_dataset = raw_datasets["test_matched" if data_args.task_name == "mnli" else "test"] + if data_args.max_predict_samples is not None: + max_predict_samples = min(len(predict_dataset), data_args.max_predict_samples) + predict_dataset = predict_dataset.select(range(max_predict_samples)) + + # Log a few random samples from the training set: + if training_args.do_train: + for index in random.sample(range(len(train_dataset)), 3): + logger.info(f"Sample {index} of the training set: {train_dataset[index]}.") + + # Get the metric function + if data_args.task_name is not None: + metric = evaluate.load("glue", data_args.task_name) + else: + metric = evaluate.load("accuracy") + + # You can define your custom compute_metrics function. It takes an `EvalPrediction` object (a namedtuple with a + # predictions and label_ids field) and has to return a dictionary string to float. + def compute_metrics(p: EvalPrediction): + preds = p.predictions[0] if isinstance(p.predictions, tuple) else p.predictions + preds = np.squeeze(preds) if is_regression else np.argmax(preds, axis=1) + if data_args.task_name is not None: + result = metric.compute(predictions=preds, references=p.label_ids) + if len(result) > 1: + result["combined_score"] = np.mean(list(result.values())).item() + return result + elif is_regression: + return {"mse": ((preds - p.label_ids) ** 2).mean().item()} + else: + return {"accuracy": (preds == p.label_ids).astype(np.float32).mean().item()} + + # Data collator will default to DataCollatorWithPadding when the tokenizer is passed to Trainer, so we change it if + # we already did the padding. + if data_args.pad_to_max_length: + data_collator = default_data_collator + elif training_args.fp16: + data_collator = DataCollatorWithPadding(tokenizer, pad_to_multiple_of=8) + else: + data_collator = None + + # Initialize our Trainer + trainer = Trainer( + model=model, + args=training_args, + train_dataset=train_dataset if training_args.do_train else None, + eval_dataset=eval_dataset if training_args.do_eval else None, + compute_metrics=compute_metrics, + tokenizer=tokenizer, + data_collator=data_collator, + ) + + # Training + if training_args.do_train: + checkpoint = None + if training_args.resume_from_checkpoint is not None: + checkpoint = training_args.resume_from_checkpoint + elif last_checkpoint is not None: + checkpoint = last_checkpoint + train_result = trainer.train(resume_from_checkpoint=checkpoint) + metrics = train_result.metrics + max_train_samples = ( + data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset) + ) + metrics["train_samples"] = min(max_train_samples, len(train_dataset)) + + trainer.save_model() # Saves the tokenizer too for easy upload + + trainer.log_metrics("train", metrics) + trainer.save_metrics("train", metrics) + trainer.save_state() + + # Evaluation + if training_args.do_eval: + logger.info("*** Evaluate ***") + + # Loop to handle MNLI double evaluation (matched, mis-matched) + tasks = [data_args.task_name] + eval_datasets = [eval_dataset] + if data_args.task_name == "mnli": + tasks.append("mnli-mm") + valid_mm_dataset = raw_datasets["validation_mismatched"] + if data_args.max_eval_samples is not None: + max_eval_samples = min(len(valid_mm_dataset), data_args.max_eval_samples) + valid_mm_dataset = valid_mm_dataset.select(range(max_eval_samples)) + eval_datasets.append(valid_mm_dataset) + combined = {} + + for eval_dataset, task in zip(eval_datasets, tasks): + metrics = trainer.evaluate(eval_dataset=eval_dataset) + + max_eval_samples = ( + data_args.max_eval_samples if data_args.max_eval_samples is not None else len(eval_dataset) + ) + metrics["eval_samples"] = min(max_eval_samples, len(eval_dataset)) + + if task == "mnli-mm": + metrics = {k + "_mm": v for k, v in metrics.items()} + if task is not None and "mnli" in task: + combined.update(metrics) + + trainer.log_metrics("eval", metrics) + trainer.save_metrics("eval", combined if task is not None and "mnli" in task else metrics) + + if training_args.do_predict: + logger.info("*** Predict ***") + + # Loop to handle MNLI double evaluation (matched, mis-matched) + tasks = [data_args.task_name] + predict_datasets = [predict_dataset] + if data_args.task_name == "mnli": + tasks.append("mnli-mm") + predict_datasets.append(raw_datasets["test_mismatched"]) + + for predict_dataset, task in zip(predict_datasets, tasks): + # Removing the `label` columns because it contains -1 and Trainer won't like that. + predict_dataset = predict_dataset.remove_columns("label") + predictions = trainer.predict(predict_dataset, metric_key_prefix="predict").predictions + predictions = np.squeeze(predictions) if is_regression else np.argmax(predictions, axis=1) + + output_predict_file = os.path.join(training_args.output_dir, f"predict_results_{task}.txt") + if trainer.is_world_process_zero(): + with open(output_predict_file, "w") as writer: + logger.info(f"***** Predict results {task} *****") + writer.write("index\tprediction\n") + for index, item in enumerate(predictions): + if is_regression: + writer.write(f"{index}\t{item:3.3f}\n") + else: + item = label_list[item] + writer.write(f"{index}\t{item}\n") + + kwargs = {"finetuned_from": model_args.model_name_or_path, "tasks": "text-classification"} + if data_args.task_name is not None: + kwargs["language"] = "en" + kwargs["dataset_tags"] = "glue" + kwargs["dataset_args"] = data_args.task_name + kwargs["dataset"] = f"GLUE {data_args.task_name.upper()}" + + if training_args.push_to_hub: + trainer.push_to_hub(**kwargs) + else: + trainer.create_model_card(**kwargs) + + +def _mp_fn(index): + # For xla_spawn (TPUs) + main() + + +if __name__ == "__main__": + main() \ No newline at end of file diff --git a/nlp/text_classification/train.sh b/nlp/text_classification/train.sh new file mode 100644 index 00000000..7f785853 --- /dev/null +++ b/nlp/text_classification/train.sh @@ -0,0 +1,12 @@ +export TASK_NAME=WNLI + +python3 run_glue.py \ + --model_name_or_path bert-base-cased \ + --task_name $TASK_NAME \ + --do_train \ + --do_eval \ + --max_seq_length 128 \ + --per_device_train_batch_size 32 \ + --learning_rate 2e-5 \ + --num_train_epochs 5 \ + --output_dir /tmp/$TASK_NAME/ \ No newline at end of file diff --git a/nlp/text_classification/train_dist.sh b/nlp/text_classification/train_dist.sh new file mode 100644 index 00000000..36874eb5 --- /dev/null +++ b/nlp/text_classification/train_dist.sh @@ -0,0 +1,12 @@ +export TASK_NAME=WNLI +python3 -m torch.distributed.launch --nproc_per_node=8 --master_port 12333 \ + run_glue.py \ + --model_name_or_path bert-base-cased \ + --task_name $TASK_NAME \ + --do_train \ + --do_eval \ + --max_seq_length 128 \ + --per_device_train_batch_size 32 \ + --learning_rate 2e-5 \ + --num_train_epochs 5 \ + --output_dir /tmp/$TASK_NAME/ \ No newline at end of file -- Gitee From 26077463646094aad72b82f052d1226f9968506b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E7=89=9B=E5=86=A0=E5=8D=9A?= Date: Fri, 24 Feb 2023 23:20:05 +0800 Subject: [PATCH 4/4] adjust directory structure --- .../bert/pytorch/README.md | 37 ++ .../bert/pytorch/requirements.txt | 10 + .../bert/pytorch/run_glue.py | 622 ++++++++++++++++++ nlp/text_classification/bert/pytorch/train.sh | 12 + .../bert/pytorch/train_dist.sh | 12 + 5 files changed, 693 insertions(+) create mode 100644 nlp/text_classification/bert/pytorch/README.md create mode 100644 nlp/text_classification/bert/pytorch/requirements.txt create mode 100644 nlp/text_classification/bert/pytorch/run_glue.py create mode 100644 nlp/text_classification/bert/pytorch/train.sh create mode 100644 nlp/text_classification/bert/pytorch/train_dist.sh diff --git a/nlp/text_classification/bert/pytorch/README.md b/nlp/text_classification/bert/pytorch/README.md new file mode 100644 index 00000000..c82b996d --- /dev/null +++ b/nlp/text_classification/bert/pytorch/README.md @@ -0,0 +1,37 @@ +# Text Classification + +# Bert-base WNLI + +## Model description + +Bert-base WNLI task Fine-tuning + +## Step 1: Installing packages + +``` shell +cd /nlp/text_classification/bert/pytorch +pip3 install -r requirements.txt +``` + +## Step 2: Training + +### On single GPU + +``` shell +bash train.sh +``` + +### Multiple GPUs on one machine + +```shell +bash train_dist.sh +``` +## Results on BI-V100 + +| GPUs | Samples/s | Loss | +|------|-----------|------| +| 1x1 | 144.5 | 0.74 | +| 1x8 | 322.74 | 0.71 | + +## Reference +https://github.com/huggingface/ diff --git a/nlp/text_classification/bert/pytorch/requirements.txt b/nlp/text_classification/bert/pytorch/requirements.txt new file mode 100644 index 00000000..dbfdb439 --- /dev/null +++ b/nlp/text_classification/bert/pytorch/requirements.txt @@ -0,0 +1,10 @@ +accelerate >= 0.12.0 +datasets >= 1.8.0 +sentencepiece != 0.1.92 +scipy +scikit-learn +protobuf +# torch >= 1.3 +evaluate +numpy == 1.23.4 +git+https://github.com/huggingface/transformers \ No newline at end of file diff --git a/nlp/text_classification/bert/pytorch/run_glue.py b/nlp/text_classification/bert/pytorch/run_glue.py new file mode 100644 index 00000000..e8ede80e --- /dev/null +++ b/nlp/text_classification/bert/pytorch/run_glue.py @@ -0,0 +1,622 @@ +#!/usr/bin/env python +# coding=utf-8 +# Copyright 2020 The HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" Finetuning the library models for sequence classification on GLUE.""" +# You can also adapt this script on your own text classification task. Pointers for this are left as comments. + +import logging +import os +import random +import sys +from dataclasses import dataclass, field +from typing import Optional + +import datasets +import evaluate +import numpy as np +from datasets import load_dataset + +import transformers +from transformers import ( + AutoConfig, + AutoModelForSequenceClassification, + AutoTokenizer, + DataCollatorWithPadding, + EvalPrediction, + HfArgumentParser, + PretrainedConfig, + Trainer, + TrainingArguments, + default_data_collator, + set_seed, +) +from transformers.trainer_utils import get_last_checkpoint +from transformers.utils import check_min_version, send_example_telemetry +from transformers.utils.versions import require_version + + +# Will error if the minimal version of Transformers is not installed. Remove at your own risks. +check_min_version("4.27.0.dev0") + +require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/text-classification/requirements.txt") + +task_to_keys = { + "cola": ("sentence", None), + "mnli": ("premise", "hypothesis"), + "mrpc": ("sentence1", "sentence2"), + "qnli": ("question", "sentence"), + "qqp": ("question1", "question2"), + "rte": ("sentence1", "sentence2"), + "sst2": ("sentence", None), + "stsb": ("sentence1", "sentence2"), + "wnli": ("sentence1", "sentence2"), +} + +logger = logging.getLogger(__name__) + + +@dataclass +class DataTrainingArguments: + """ + Arguments pertaining to what data we are going to input our model for training and eval. + + Using `HfArgumentParser` we can turn this class + into argparse arguments to be able to specify them on + the command line. + """ + + task_name: Optional[str] = field( + default=None, + metadata={"help": "The name of the task to train on: " + ", ".join(task_to_keys.keys())}, + ) + dataset_name: Optional[str] = field( + default=None, metadata={"help": "The name of the dataset to use (via the datasets library)."} + ) + dataset_config_name: Optional[str] = field( + default=None, metadata={"help": "The configuration name of the dataset to use (via the datasets library)."} + ) + max_seq_length: int = field( + default=128, + metadata={ + "help": ( + "The maximum total input sequence length after tokenization. Sequences longer " + "than this will be truncated, sequences shorter will be padded." + ) + }, + ) + overwrite_cache: bool = field( + default=False, metadata={"help": "Overwrite the cached preprocessed datasets or not."} + ) + pad_to_max_length: bool = field( + default=True, + metadata={ + "help": ( + "Whether to pad all samples to `max_seq_length`. " + "If False, will pad the samples dynamically when batching to the maximum length in the batch." + ) + }, + ) + max_train_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of training examples to this " + "value if set." + ) + }, + ) + max_eval_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of evaluation examples to this " + "value if set." + ) + }, + ) + max_predict_samples: Optional[int] = field( + default=None, + metadata={ + "help": ( + "For debugging purposes or quicker training, truncate the number of prediction examples to this " + "value if set." + ) + }, + ) + train_file: Optional[str] = field( + default=None, metadata={"help": "A csv or a json file containing the training data."} + ) + validation_file: Optional[str] = field( + default=None, metadata={"help": "A csv or a json file containing the validation data."} + ) + test_file: Optional[str] = field(default=None, metadata={"help": "A csv or a json file containing the test data."}) + + def __post_init__(self): + if self.task_name is not None: + self.task_name = self.task_name.lower() + if self.task_name not in task_to_keys.keys(): + raise ValueError("Unknown task, you should pick one in " + ",".join(task_to_keys.keys())) + elif self.dataset_name is not None: + pass + elif self.train_file is None or self.validation_file is None: + raise ValueError("Need either a GLUE task, a training/validation file or a dataset name.") + else: + train_extension = self.train_file.split(".")[-1] + assert train_extension in ["csv", "json"], "`train_file` should be a csv or a json file." + validation_extension = self.validation_file.split(".")[-1] + assert ( + validation_extension == train_extension + ), "`validation_file` should have the same extension (csv or json) as `train_file`." + + +@dataclass +class ModelArguments: + """ + Arguments pertaining to which model/config/tokenizer we are going to fine-tune from. + """ + + model_name_or_path: str = field( + metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"} + ) + config_name: Optional[str] = field( + default=None, metadata={"help": "Pretrained config name or path if not the same as model_name"} + ) + tokenizer_name: Optional[str] = field( + default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"} + ) + cache_dir: Optional[str] = field( + default=None, + metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}, + ) + use_fast_tokenizer: bool = field( + default=True, + metadata={"help": "Whether to use one of the fast tokenizer (backed by the tokenizers library) or not."}, + ) + model_revision: str = field( + default="main", + metadata={"help": "The specific model version to use (can be a branch name, tag name or commit id)."}, + ) + use_auth_token: bool = field( + default=False, + metadata={ + "help": ( + "Will use the token generated when running `huggingface-cli login` (necessary to use this script " + "with private models)." + ) + }, + ) + ignore_mismatched_sizes: bool = field( + default=False, + metadata={"help": "Will enable to load a pretrained model whose head dimensions are different."}, + ) + + +def main(): + # See all possible arguments in src/transformers/training_args.py + # or by passing the --help flag to this script. + # We now keep distinct sets of args, for a cleaner separation of concerns. + + parser = HfArgumentParser((ModelArguments, DataTrainingArguments, TrainingArguments)) + if len(sys.argv) == 2 and sys.argv[1].endswith(".json"): + # If we pass only one argument to the script and it's the path to a json file, + # let's parse it to get our arguments. + model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1])) + else: + model_args, data_args, training_args = parser.parse_args_into_dataclasses() + + # 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. + send_example_telemetry("run_glue", 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 = training_args.get_process_log_level() + logger.setLevel(log_level) + datasets.utils.logging.set_verbosity(log_level) + transformers.utils.logging.set_verbosity(log_level) + transformers.utils.logging.enable_default_handler() + transformers.utils.logging.enable_explicit_format() + + # Log on each process the small summary: + logger.warning( + f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}" + + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}" + ) + logger.info(f"Training/evaluation parameters {training_args}") + + # Detecting last checkpoint. + last_checkpoint = None + if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir: + last_checkpoint = get_last_checkpoint(training_args.output_dir) + if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0: + raise ValueError( + f"Output directory ({training_args.output_dir}) already exists and is not empty. " + "Use --overwrite_output_dir to overcome." + ) + elif last_checkpoint is not None and training_args.resume_from_checkpoint is None: + logger.info( + f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change " + "the `--output_dir` or add `--overwrite_output_dir` to train from scratch." + ) + + # Set seed before initializing model. + set_seed(training_args.seed) + + # Get the datasets: you can either provide your own CSV/JSON training and evaluation files (see below) + # or specify a GLUE benchmark task (the dataset will be downloaded automatically from the datasets Hub). + # + # For CSV/JSON files, this script will use as labels the column called 'label' and as pair of sentences the + # sentences in columns called 'sentence1' and 'sentence2' if such column exists or the first two columns not named + # label if at least two columns are provided. + # + # If the CSVs/JSONs contain only one non-label column, the script does single sentence classification on this + # single column. You can easily tweak this behavior (see below) + # + # In distributed training, the load_dataset function guarantee that only one local process can concurrently + # download the dataset. + if data_args.task_name is not None: + # Downloading and loading a dataset from the hub. + raw_datasets = load_dataset( + "glue", + data_args.task_name, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + elif data_args.dataset_name is not None: + # Downloading and loading a dataset from the hub. + raw_datasets = load_dataset( + data_args.dataset_name, + data_args.dataset_config_name, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + else: + # Loading a dataset from your local files. + # CSV/JSON training and evaluation files are needed. + data_files = {"train": data_args.train_file, "validation": data_args.validation_file} + + # Get the test dataset: you can provide your own CSV/JSON test file (see below) + # when you use `do_predict` without specifying a GLUE benchmark task. + if training_args.do_predict: + if data_args.test_file is not None: + train_extension = data_args.train_file.split(".")[-1] + test_extension = data_args.test_file.split(".")[-1] + assert ( + test_extension == train_extension + ), "`test_file` should have the same extension (csv or json) as `train_file`." + data_files["test"] = data_args.test_file + else: + raise ValueError("Need either a GLUE task or a test file for `do_predict`.") + + for key in data_files.keys(): + logger.info(f"load a local file for {key}: {data_files[key]}") + + if data_args.train_file.endswith(".csv"): + # Loading a dataset from local csv files + raw_datasets = load_dataset( + "csv", + data_files=data_files, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + else: + # Loading a dataset from local json files + raw_datasets = load_dataset( + "json", + data_files=data_files, + cache_dir=model_args.cache_dir, + use_auth_token=True if model_args.use_auth_token else None, + ) + # See more about loading any type of standard or custom dataset at + # https://huggingface.co/docs/datasets/loading_datasets.html. + + # Labels + if data_args.task_name is not None: + is_regression = data_args.task_name == "stsb" + if not is_regression: + label_list = raw_datasets["train"].features["label"].names + num_labels = len(label_list) + else: + num_labels = 1 + else: + # Trying to have good defaults here, don't hesitate to tweak to your needs. + is_regression = raw_datasets["train"].features["label"].dtype in ["float32", "float64"] + if is_regression: + num_labels = 1 + else: + # A useful fast method: + # https://huggingface.co/docs/datasets/package_reference/main_classes.html#datasets.Dataset.unique + label_list = raw_datasets["train"].unique("label") + label_list.sort() # Let's sort it for determinism + num_labels = len(label_list) + + # Load pretrained model and tokenizer + # + # In distributed training, the .from_pretrained methods guarantee that only one local process can concurrently + # download model & vocab. + config = AutoConfig.from_pretrained( + model_args.config_name if model_args.config_name else model_args.model_name_or_path, + num_labels=num_labels, + finetuning_task=data_args.task_name, + cache_dir=model_args.cache_dir, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ) + tokenizer = AutoTokenizer.from_pretrained( + model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path, + cache_dir=model_args.cache_dir, + use_fast=model_args.use_fast_tokenizer, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ) + model = AutoModelForSequenceClassification.from_pretrained( + model_args.model_name_or_path, + from_tf=bool(".ckpt" in model_args.model_name_or_path), + config=config, + cache_dir=model_args.cache_dir, + revision=model_args.model_revision, + use_auth_token=True if model_args.use_auth_token else None, + ignore_mismatched_sizes=model_args.ignore_mismatched_sizes, + ) + + # Preprocessing the raw_datasets + if data_args.task_name is not None: + sentence1_key, sentence2_key = task_to_keys[data_args.task_name] + else: + # Again, we try to have some nice defaults but don't hesitate to tweak to your use case. + non_label_column_names = [name for name in raw_datasets["train"].column_names if name != "label"] + if "sentence1" in non_label_column_names and "sentence2" in non_label_column_names: + sentence1_key, sentence2_key = "sentence1", "sentence2" + else: + if len(non_label_column_names) >= 2: + sentence1_key, sentence2_key = non_label_column_names[:2] + else: + sentence1_key, sentence2_key = non_label_column_names[0], None + + # Padding strategy + if data_args.pad_to_max_length: + padding = "max_length" + else: + # We will pad later, dynamically at batch creation, to the max sequence length in each batch + padding = False + + # Some models have set the order of the labels to use, so let's make sure we do use it. + label_to_id = None + if ( + model.config.label2id != PretrainedConfig(num_labels=num_labels).label2id + and data_args.task_name is not None + and not is_regression + ): + # Some have all caps in their config, some don't. + label_name_to_id = {k.lower(): v for k, v in model.config.label2id.items()} + if list(sorted(label_name_to_id.keys())) == list(sorted(label_list)): + label_to_id = {i: int(label_name_to_id[label_list[i]]) for i in range(num_labels)} + else: + logger.warning( + "Your model seems to have been trained with labels, but they don't match the dataset: ", + f"model labels: {list(sorted(label_name_to_id.keys()))}, dataset labels: {list(sorted(label_list))}." + "\nIgnoring the model labels as a result.", + ) + elif data_args.task_name is None and not is_regression: + label_to_id = {v: i for i, v in enumerate(label_list)} + + if label_to_id is not None: + model.config.label2id = label_to_id + model.config.id2label = {id: label for label, id in config.label2id.items()} + elif data_args.task_name is not None and not is_regression: + model.config.label2id = {l: i for i, l in enumerate(label_list)} + model.config.id2label = {id: label for label, id in config.label2id.items()} + + if data_args.max_seq_length > tokenizer.model_max_length: + logger.warning( + f"The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the" + f"model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}." + ) + max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length) + + def preprocess_function(examples): + # Tokenize the texts + args = ( + (examples[sentence1_key],) if sentence2_key is None else (examples[sentence1_key], examples[sentence2_key]) + ) + result = tokenizer(*args, padding=padding, max_length=max_seq_length, truncation=True) + + # Map labels to IDs (not necessary for GLUE tasks) + if label_to_id is not None and "label" in examples: + result["label"] = [(label_to_id[l] if l != -1 else -1) for l in examples["label"]] + return result + + with training_args.main_process_first(desc="dataset map pre-processing"): + raw_datasets = raw_datasets.map( + preprocess_function, + batched=True, + load_from_cache_file=not data_args.overwrite_cache, + desc="Running tokenizer on dataset", + ) + if training_args.do_train: + if "train" not in raw_datasets: + raise ValueError("--do_train requires a train dataset") + train_dataset = raw_datasets["train"] + if data_args.max_train_samples is not None: + max_train_samples = min(len(train_dataset), data_args.max_train_samples) + train_dataset = train_dataset.select(range(max_train_samples)) + + if training_args.do_eval: + if "validation" not in raw_datasets and "validation_matched" not in raw_datasets: + raise ValueError("--do_eval requires a validation dataset") + eval_dataset = raw_datasets["validation_matched" if data_args.task_name == "mnli" else "validation"] + if data_args.max_eval_samples is not None: + max_eval_samples = min(len(eval_dataset), data_args.max_eval_samples) + eval_dataset = eval_dataset.select(range(max_eval_samples)) + + if training_args.do_predict or data_args.task_name is not None or data_args.test_file is not None: + if "test" not in raw_datasets and "test_matched" not in raw_datasets: + raise ValueError("--do_predict requires a test dataset") + predict_dataset = raw_datasets["test_matched" if data_args.task_name == "mnli" else "test"] + if data_args.max_predict_samples is not None: + max_predict_samples = min(len(predict_dataset), data_args.max_predict_samples) + predict_dataset = predict_dataset.select(range(max_predict_samples)) + + # Log a few random samples from the training set: + if training_args.do_train: + for index in random.sample(range(len(train_dataset)), 3): + logger.info(f"Sample {index} of the training set: {train_dataset[index]}.") + + # Get the metric function + if data_args.task_name is not None: + metric = evaluate.load("glue", data_args.task_name) + else: + metric = evaluate.load("accuracy") + + # You can define your custom compute_metrics function. It takes an `EvalPrediction` object (a namedtuple with a + # predictions and label_ids field) and has to return a dictionary string to float. + def compute_metrics(p: EvalPrediction): + preds = p.predictions[0] if isinstance(p.predictions, tuple) else p.predictions + preds = np.squeeze(preds) if is_regression else np.argmax(preds, axis=1) + if data_args.task_name is not None: + result = metric.compute(predictions=preds, references=p.label_ids) + if len(result) > 1: + result["combined_score"] = np.mean(list(result.values())).item() + return result + elif is_regression: + return {"mse": ((preds - p.label_ids) ** 2).mean().item()} + else: + return {"accuracy": (preds == p.label_ids).astype(np.float32).mean().item()} + + # Data collator will default to DataCollatorWithPadding when the tokenizer is passed to Trainer, so we change it if + # we already did the padding. + if data_args.pad_to_max_length: + data_collator = default_data_collator + elif training_args.fp16: + data_collator = DataCollatorWithPadding(tokenizer, pad_to_multiple_of=8) + else: + data_collator = None + + # Initialize our Trainer + trainer = Trainer( + model=model, + args=training_args, + train_dataset=train_dataset if training_args.do_train else None, + eval_dataset=eval_dataset if training_args.do_eval else None, + compute_metrics=compute_metrics, + tokenizer=tokenizer, + data_collator=data_collator, + ) + + # Training + if training_args.do_train: + checkpoint = None + if training_args.resume_from_checkpoint is not None: + checkpoint = training_args.resume_from_checkpoint + elif last_checkpoint is not None: + checkpoint = last_checkpoint + train_result = trainer.train(resume_from_checkpoint=checkpoint) + metrics = train_result.metrics + max_train_samples = ( + data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset) + ) + metrics["train_samples"] = min(max_train_samples, len(train_dataset)) + + trainer.save_model() # Saves the tokenizer too for easy upload + + trainer.log_metrics("train", metrics) + trainer.save_metrics("train", metrics) + trainer.save_state() + + # Evaluation + if training_args.do_eval: + logger.info("*** Evaluate ***") + + # Loop to handle MNLI double evaluation (matched, mis-matched) + tasks = [data_args.task_name] + eval_datasets = [eval_dataset] + if data_args.task_name == "mnli": + tasks.append("mnli-mm") + valid_mm_dataset = raw_datasets["validation_mismatched"] + if data_args.max_eval_samples is not None: + max_eval_samples = min(len(valid_mm_dataset), data_args.max_eval_samples) + valid_mm_dataset = valid_mm_dataset.select(range(max_eval_samples)) + eval_datasets.append(valid_mm_dataset) + combined = {} + + for eval_dataset, task in zip(eval_datasets, tasks): + metrics = trainer.evaluate(eval_dataset=eval_dataset) + + max_eval_samples = ( + data_args.max_eval_samples if data_args.max_eval_samples is not None else len(eval_dataset) + ) + metrics["eval_samples"] = min(max_eval_samples, len(eval_dataset)) + + if task == "mnli-mm": + metrics = {k + "_mm": v for k, v in metrics.items()} + if task is not None and "mnli" in task: + combined.update(metrics) + + trainer.log_metrics("eval", metrics) + trainer.save_metrics("eval", combined if task is not None and "mnli" in task else metrics) + + if training_args.do_predict: + logger.info("*** Predict ***") + + # Loop to handle MNLI double evaluation (matched, mis-matched) + tasks = [data_args.task_name] + predict_datasets = [predict_dataset] + if data_args.task_name == "mnli": + tasks.append("mnli-mm") + predict_datasets.append(raw_datasets["test_mismatched"]) + + for predict_dataset, task in zip(predict_datasets, tasks): + # Removing the `label` columns because it contains -1 and Trainer won't like that. + predict_dataset = predict_dataset.remove_columns("label") + predictions = trainer.predict(predict_dataset, metric_key_prefix="predict").predictions + predictions = np.squeeze(predictions) if is_regression else np.argmax(predictions, axis=1) + + output_predict_file = os.path.join(training_args.output_dir, f"predict_results_{task}.txt") + if trainer.is_world_process_zero(): + with open(output_predict_file, "w") as writer: + logger.info(f"***** Predict results {task} *****") + writer.write("index\tprediction\n") + for index, item in enumerate(predictions): + if is_regression: + writer.write(f"{index}\t{item:3.3f}\n") + else: + item = label_list[item] + writer.write(f"{index}\t{item}\n") + + kwargs = {"finetuned_from": model_args.model_name_or_path, "tasks": "text-classification"} + if data_args.task_name is not None: + kwargs["language"] = "en" + kwargs["dataset_tags"] = "glue" + kwargs["dataset_args"] = data_args.task_name + kwargs["dataset"] = f"GLUE {data_args.task_name.upper()}" + + if training_args.push_to_hub: + trainer.push_to_hub(**kwargs) + else: + trainer.create_model_card(**kwargs) + + +def _mp_fn(index): + # For xla_spawn (TPUs) + main() + + +if __name__ == "__main__": + main() \ No newline at end of file diff --git a/nlp/text_classification/bert/pytorch/train.sh b/nlp/text_classification/bert/pytorch/train.sh new file mode 100644 index 00000000..7f785853 --- /dev/null +++ b/nlp/text_classification/bert/pytorch/train.sh @@ -0,0 +1,12 @@ +export TASK_NAME=WNLI + +python3 run_glue.py \ + --model_name_or_path bert-base-cased \ + --task_name $TASK_NAME \ + --do_train \ + --do_eval \ + --max_seq_length 128 \ + --per_device_train_batch_size 32 \ + --learning_rate 2e-5 \ + --num_train_epochs 5 \ + --output_dir /tmp/$TASK_NAME/ \ No newline at end of file diff --git a/nlp/text_classification/bert/pytorch/train_dist.sh b/nlp/text_classification/bert/pytorch/train_dist.sh new file mode 100644 index 00000000..36874eb5 --- /dev/null +++ b/nlp/text_classification/bert/pytorch/train_dist.sh @@ -0,0 +1,12 @@ +export TASK_NAME=WNLI +python3 -m torch.distributed.launch --nproc_per_node=8 --master_port 12333 \ + run_glue.py \ + --model_name_or_path bert-base-cased \ + --task_name $TASK_NAME \ + --do_train \ + --do_eval \ + --max_seq_length 128 \ + --per_device_train_batch_size 32 \ + --learning_rate 2e-5 \ + --num_train_epochs 5 \ + --output_dir /tmp/$TASK_NAME/ \ No newline at end of file -- Gitee