# coding=utf-8 # Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team. # Copyright (c) 2018, NVIDIA CORPORATION. All rights reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. """ Fine-tuning the library models for language modeling on a text file (GPT, GPT-2, BERT, RoBERTa). GPT and GPT-2 are fine-tuned using a causal language modeling (CLM) loss while BERT and RoBERTa are fine-tuned using a masked language modeling (MLM) loss. """ from __future__ import absolute_import, division, print_function import argparse import glob import logging import os import pickle import random import re import shutil import numpy as np import torch from torch.utils.data import DataLoader, Dataset, SequentialSampler, RandomSampler,TensorDataset from torch.utils.data.distributed import DistributedSampler import json try: from torch.utils.tensorboard import SummaryWriter except: from tensorboardX import SummaryWriter from tqdm import tqdm, trange import multiprocessing from model import Model cpu_cont = multiprocessing.cpu_count() from transformers import (WEIGHTS_NAME, AdamW, get_linear_schedule_with_warmup, BertConfig, BertForMaskedLM, BertTokenizer, BertForSequenceClassification, GPT2Config, GPT2LMHeadModel, GPT2Tokenizer, OpenAIGPTConfig, OpenAIGPTLMHeadModel, OpenAIGPTTokenizer, RobertaConfig, RobertaForSequenceClassification, RobertaTokenizer, DistilBertConfig, DistilBertForMaskedLM, DistilBertForSequenceClassification, DistilBertTokenizer) logger = logging.getLogger(__name__) MODEL_CLASSES = { 'gpt2': (GPT2Config, GPT2LMHeadModel, GPT2Tokenizer), 'openai-gpt': (OpenAIGPTConfig, OpenAIGPTLMHeadModel, OpenAIGPTTokenizer), 'bert': (BertConfig, BertForSequenceClassification, BertTokenizer), 'roberta': (RobertaConfig, RobertaForSequenceClassification, RobertaTokenizer), 'distilbert': (DistilBertConfig, DistilBertForSequenceClassification, DistilBertTokenizer) } class InputFeatures(object): """A single training/test features for a example.""" def __init__(self, input_tokens, input_ids, idx, label, ): self.input_tokens = input_tokens self.input_ids = input_ids self.idx=str(idx) self.label=label def convert_examples_to_features(js,tokenizer,args): #source code=' '.join(js['func'].split()) code_tokens=tokenizer.tokenize(code)[:args.block_size-2] source_tokens =[tokenizer.cls_token]+code_tokens+[tokenizer.sep_token] source_ids = tokenizer.convert_tokens_to_ids(source_tokens) padding_length = args.block_size - len(source_ids) source_ids+=[tokenizer.pad_token_id]*padding_length return InputFeatures(source_tokens,source_ids,js['idx'],js['target']) class TextDataset(Dataset): def __init__(self, tokenizer, args, file_path=None): self.examples = [] with open(file_path) as f: for line in f: js=json.loads(line.strip()) self.examples.append(convert_examples_to_features(js,tokenizer,args)) if 'train' in file_path: for idx, example in enumerate(self.examples[:3]): logger.info("*** Example ***") logger.info("idx: {}".format(idx)) logger.info("label: {}".format(example.label)) logger.info("input_tokens: {}".format([x.replace('\u0120','_') for x in example.input_tokens])) logger.info("input_ids: {}".format(' '.join(map(str, example.input_ids)))) def __len__(self): return len(self.examples) def __getitem__(self, i): return torch.tensor(self.examples[i].input_ids),torch.tensor(self.examples[i].label) def set_seed(seed=42): random.seed(seed) os.environ['PYHTONHASHSEED'] = str(seed) np.random.seed(seed) torch.manual_seed(seed) torch.cuda.manual_seed(seed) torch.backends.cudnn.deterministic = True def train(args, train_dataset, model, tokenizer): """ Train the model """ args.train_batch_size = args.per_gpu_train_batch_size * max(1, args.n_gpu) train_sampler = RandomSampler(train_dataset) if args.local_rank == -1 else DistributedSampler(train_dataset) train_dataloader = DataLoader(train_dataset, sampler=train_sampler, batch_size=args.train_batch_size,num_workers=4,pin_memory=True) args.max_steps=args.epoch*len( train_dataloader) args.save_steps=len( train_dataloader) args.warmup_steps=len( train_dataloader) args.logging_steps=len( train_dataloader) args.num_train_epochs=args.epoch model.to(args.device) # Prepare optimizer and schedule (linear warmup and decay) no_decay = ['bias', 'LayerNorm.weight'] optimizer_grouped_parameters = [ {'params': [p for n, p in model.named_parameters() if not any(nd in n for nd in no_decay)], 'weight_decay': args.weight_decay}, {'params': [p for n, p in model.named_parameters() if any(nd in n for nd in no_decay)], 'weight_decay': 0.0} ] optimizer = AdamW(optimizer_grouped_parameters, lr=args.learning_rate, eps=args.adam_epsilon) scheduler = get_linear_schedule_with_warmup(optimizer, num_warmup_steps=args.max_steps*0.1, num_training_steps=args.max_steps) if args.fp16: try: from apex import amp except ImportError: raise ImportError("Please install apex from https://www.github.com/nvidia/apex to use fp16 training.") model, optimizer = amp.initialize(model, optimizer, opt_level=args.fp16_opt_level) # multi-gpu training (should be after apex fp16 initialization) if args.n_gpu > 1: model = torch.nn.DataParallel(model) # Distributed training (should be after apex fp16 initialization) if args.local_rank != -1: model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.local_rank], output_device=args.local_rank, find_unused_parameters=True) checkpoint_last = os.path.join(args.output_dir, 'checkpoint-last') scheduler_last = os.path.join(checkpoint_last, 'scheduler.pt') optimizer_last = os.path.join(checkpoint_last, 'optimizer.pt') if os.path.exists(scheduler_last): scheduler.load_state_dict(torch.load(scheduler_last)) if os.path.exists(optimizer_last): optimizer.load_state_dict(torch.load(optimizer_last)) # Train! logger.info("***** Running training *****") logger.info(" Num examples = %d", len(train_dataset)) logger.info(" Num Epochs = %d", args.num_train_epochs) logger.info(" Instantaneous batch size per GPU = %d", args.per_gpu_train_batch_size) logger.info(" Total train batch size (w. parallel, distributed & accumulation) = %d", args.train_batch_size * args.gradient_accumulation_steps * ( torch.distributed.get_world_size() if args.local_rank != -1 else 1)) logger.info(" Gradient Accumulation steps = %d", args.gradient_accumulation_steps) logger.info(" Total optimization steps = %d", args.max_steps) global_step = args.start_step tr_loss, logging_loss,avg_loss,tr_nb,tr_num,train_loss = 0.0, 0.0,0.0,0,0,0 best_mrr=0.0 best_acc=0.0 # model.resize_token_embeddings(len(tokenizer)) model.zero_grad() # Initialize early stopping parameters at the start of training early_stopping_counter = 0 best_loss = None for idx in range(args.start_epoch, int(args.num_train_epochs)): bar = tqdm(train_dataloader,total=len(train_dataloader)) tr_num=0 train_loss=0 for step, batch in enumerate(bar): inputs = batch[0].to(args.device) labels=batch[1].to(args.device) model.train() loss,logits = model(inputs,labels) if args.n_gpu > 1: loss = loss.mean() # mean() to average on multi-gpu parallel training if args.gradient_accumulation_steps > 1: loss = loss / args.gradient_accumulation_steps if args.fp16: with amp.scale_loss(loss, optimizer) as scaled_loss: scaled_loss.backward() torch.nn.utils.clip_grad_norm_(amp.master_params(optimizer), args.max_grad_norm) else: loss.backward() torch.nn.utils.clip_grad_norm_(model.parameters(), args.max_grad_norm) tr_loss += loss.item() tr_num+=1 train_loss+=loss.item() if avg_loss==0: avg_loss=tr_loss avg_loss=round(train_loss/tr_num,5) bar.set_description("epoch {} loss {}".format(idx,avg_loss)) if (step + 1) % args.gradient_accumulation_steps == 0: optimizer.step() optimizer.zero_grad() scheduler.step() global_step += 1 output_flag=True avg_loss=round(np.exp((tr_loss - logging_loss) /(global_step- tr_nb)),4) if args.local_rank in [-1, 0] and args.logging_steps > 0 and global_step % args.logging_steps == 0: logging_loss = tr_loss tr_nb=global_step if args.local_rank in [-1, 0] and args.save_steps > 0 and global_step % args.save_steps == 0: if args.local_rank == -1 and args.evaluate_during_training: # Only evaluate when single GPU otherwise metrics may not average well results = evaluate(args, model, tokenizer,eval_when_training=True) for key, value in results.items(): logger.info(" %s = %s", key, round(value,4)) # Save model checkpoint if results['eval_acc']>best_acc: best_acc=results['eval_acc'] logger.info(" "+"*"*20) logger.info(" Best acc:%s",round(best_acc,4)) logger.info(" "+"*"*20) checkpoint_prefix = 'checkpoint-best-acc' output_dir = os.path.join(args.output_dir, '{}'.format(checkpoint_prefix)) if not os.path.exists(output_dir): os.makedirs(output_dir) model_to_save = model.module if hasattr(model,'module') else model output_dir = os.path.join(output_dir, '{}'.format('model.bin')) torch.save(model_to_save.state_dict(), output_dir) logger.info("Saving model checkpoint to %s", output_dir) # 每一轮记录checkpoint output_dir = os.path.join(args.output_dir, 'epoch_{}'.format(idx+1)) if not os.path.exists(output_dir): os.makedirs(output_dir) model_to_save = model.module if hasattr(model, 'module') else model ckpt_output_path = os.path.join(output_dir, 'subject_model.pth') logger.info("Saving model checkpoint to %s", ckpt_output_path) torch.save(model_to_save.state_dict(), ckpt_output_path) # Calculate average loss for the epoch avg_loss = train_loss / tr_num # Check for early stopping condition if args.early_stopping_patience is not None: if best_loss is None or avg_loss < best_loss - args.min_loss_delta: best_loss = avg_loss early_stopping_counter = 0 else: early_stopping_counter += 1 if early_stopping_counter >= args.early_stopping_patience: logger.info("Early stopping") break # Exit the loop early def evaluate(args, model, tokenizer,eval_when_training=False): # Loop to handle MNLI double evaluation (matched, mis-matched) eval_output_dir = args.output_dir eval_dataset = TextDataset(tokenizer, args,args.eval_data_file) if not os.path.exists(eval_output_dir) and args.local_rank in [-1, 0]: os.makedirs(eval_output_dir) args.eval_batch_size = args.per_gpu_eval_batch_size * max(1, args.n_gpu) # Note that DistributedSampler samples randomly eval_sampler = SequentialSampler(eval_dataset) if args.local_rank == -1 else DistributedSampler(eval_dataset) eval_dataloader = DataLoader(eval_dataset, sampler=eval_sampler, batch_size=args.eval_batch_size,num_workers=4,pin_memory=True) # multi-gpu evaluate if args.n_gpu > 1 and eval_when_training is False: model = torch.nn.DataParallel(model) # Eval! logger.info("***** Running evaluation *****") logger.info(" Num examples = %d", len(eval_dataset)) logger.info(" Batch size = %d", args.eval_batch_size) eval_loss = 0.0 nb_eval_steps = 0 model.eval() logits=[] labels=[] for batch in eval_dataloader: inputs = batch[0].to(args.device) label=batch[1].to(args.device) with torch.no_grad(): lm_loss,logit = model(inputs,label) eval_loss += lm_loss.mean().item() logits.append(logit.cpu().numpy()) labels.append(label.cpu().numpy()) nb_eval_steps += 1 logits=np.concatenate(logits,0) labels=np.concatenate(labels,0) preds=logits[:,0]>0.5 eval_acc=np.mean(labels==preds) eval_loss = eval_loss / nb_eval_steps perplexity = torch.tensor(eval_loss) result = { "eval_loss": float(perplexity), "eval_acc":round(eval_acc,4), } return result def test(args, model, tokenizer): # Loop to handle MNLI double evaluation (matched, mis-matched) eval_dataset = TextDataset(tokenizer, args,args.test_data_file) args.eval_batch_size = args.per_gpu_eval_batch_size * max(1, args.n_gpu) # Note that DistributedSampler samples randomly eval_sampler = SequentialSampler(eval_dataset) if args.local_rank == -1 else DistributedSampler(eval_dataset) eval_dataloader = DataLoader(eval_dataset, sampler=eval_sampler, batch_size=args.eval_batch_size) # multi-gpu evaluate if args.n_gpu > 1: model = torch.nn.DataParallel(model) # Eval! logger.info("***** Running Test *****") logger.info(" Num examples = %d", len(eval_dataset)) logger.info(" Batch size = %d", args.eval_batch_size) eval_loss = 0.0 nb_eval_steps = 0 model.eval() logits=[] labels=[] for batch in tqdm(eval_dataloader,total=len(eval_dataloader)): inputs = batch[0].to(args.device) label=batch[1].to(args.device) with torch.no_grad(): logit = model(inputs) logits.append(logit.cpu().numpy()) labels.append(label.cpu().numpy()) logits=np.concatenate(logits,0) labels=np.concatenate(labels,0) preds=logits[:,0]>0.5 with open(os.path.join(args.output_dir,"predictions.txt"),'w') as f: for example,pred in zip(eval_dataset.examples,preds): if pred: f.write(example.idx+'\t1\n') else: f.write(example.idx+'\t0\n') def main(): parser = argparse.ArgumentParser() ## Required parameters parser.add_argument("--train_data_file", default=None, type=str, required=True, help="The input training data file (a text file).") parser.add_argument("--output_dir", default=None, type=str, required=True, help="The output directory where the model predictions and checkpoints will be written.") ## Other parameters parser.add_argument("--eval_data_file", default=None, type=str, help="An optional input evaluation data file to evaluate the perplexity on (a text file).") parser.add_argument("--test_data_file", default=None, type=str, help="An optional input evaluation data file to evaluate the perplexity on (a text file).") parser.add_argument("--model_type", default="bert", type=str, help="The model architecture to be fine-tuned.") parser.add_argument("--model_name_or_path", default=None, type=str, help="The model checkpoint for weights initialization.") parser.add_argument("--mlm", action='store_true', help="Train with masked-language modeling loss instead of language modeling.") parser.add_argument("--mlm_probability", type=float, default=0.15, help="Ratio of tokens to mask for masked language modeling loss") parser.add_argument("--config_name", default="", type=str, help="Optional pretrained config name or path if not the same as model_name_or_path") parser.add_argument("--tokenizer_name", default="", type=str, help="Optional pretrained tokenizer name or path if not the same as model_name_or_path") parser.add_argument("--cache_dir", default="", type=str, help="Optional directory to store the pre-trained models downloaded from s3 (instread of the default one)") parser.add_argument("--block_size", default=-1, type=int, help="Optional input sequence length after tokenization." "The training dataset will be truncated in block of this size for training." "Default to the model max input length for single sentence inputs (take into account special tokens).") parser.add_argument("--do_train", action='store_true', help="Whether to run training.") parser.add_argument("--do_eval", action='store_true', help="Whether to run eval on the dev set.") parser.add_argument("--do_test", action='store_true', help="Whether to run eval on the dev set.") parser.add_argument("--evaluate_during_training", action='store_true', help="Run evaluation during training at each logging step.") parser.add_argument("--do_lower_case", action='store_true', help="Set this flag if you are using an uncased model.") parser.add_argument("--train_batch_size", default=4, type=int, help="Batch size per GPU/CPU for training.") parser.add_argument("--eval_batch_size", default=4, type=int, help="Batch size per GPU/CPU for evaluation.") parser.add_argument('--gradient_accumulation_steps', type=int, default=1, help="Number of updates steps to accumulate before performing a backward/update pass.") parser.add_argument("--learning_rate", default=5e-5, type=float, help="The initial learning rate for Adam.") parser.add_argument("--weight_decay", default=0.0, type=float, help="Weight deay if we apply some.") parser.add_argument("--adam_epsilon", default=1e-8, type=float, help="Epsilon for Adam optimizer.") parser.add_argument("--max_grad_norm", default=1.0, type=float, help="Max gradient norm.") parser.add_argument("--num_train_epochs", default=1.0, type=float, help="Total number of training epochs to perform.") parser.add_argument("--max_steps", default=-1, type=int, help="If > 0: set total number of training steps to perform. Override num_train_epochs.") parser.add_argument("--warmup_steps", default=0, type=int, help="Linear warmup over warmup_steps.") parser.add_argument('--logging_steps', type=int, default=50, help="Log every X updates steps.") parser.add_argument('--save_steps', type=int, default=50, help="Save checkpoint every X updates steps.") parser.add_argument('--save_total_limit', type=int, default=None, help='Limit the total amount of checkpoints, delete the older checkpoints in the output_dir, does not delete by default') parser.add_argument("--eval_all_checkpoints", action='store_true', help="Evaluate all checkpoints starting with the same prefix as model_name_or_path ending and ending with step number") parser.add_argument("--no_cuda", action='store_true', help="Avoid using CUDA when available") parser.add_argument('--overwrite_output_dir', action='store_true', help="Overwrite the content of the output directory") parser.add_argument('--overwrite_cache', action='store_true', help="Overwrite the cached training and evaluation sets") parser.add_argument('--seed', type=int, default=42, help="random seed for initialization") parser.add_argument('--epoch', type=int, default=42, help="random seed for initialization") parser.add_argument('--fp16', action='store_true', help="Whether to use 16-bit (mixed) precision (through NVIDIA apex) instead of 32-bit") parser.add_argument('--fp16_opt_level', type=str, default='O1', help="For fp16: Apex AMP optimization level selected in ['O0', 'O1', 'O2', and 'O3']." "See details at https://nvidia.github.io/apex/amp.html") parser.add_argument("--local_rank", type=int, default=-1, help="For distributed training: local_rank") parser.add_argument('--server_ip', type=str, default='', help="For distant debugging.") parser.add_argument('--server_port', type=str, default='', help="For distant debugging.") # Add early stopping parameters and dropout probability parameters parser.add_argument("--early_stopping_patience", type=int, default=None, help="Number of epochs with no improvement after which training will be stopped.") parser.add_argument("--min_loss_delta", type=float, default=0.001, help="Minimum change in the loss required to qualify as an improvement.") parser.add_argument('--dropout_probability', type=float, default=0, help='dropout probability') args = parser.parse_args() # Setup distant debugging if needed if args.server_ip and args.server_port: # Distant debugging - see https://code.visualstudio.com/docs/python/debugging#_attach-to-a-local-script import ptvsd print("Waiting for debugger attach") ptvsd.enable_attach(address=(args.server_ip, args.server_port), redirect_output=True) ptvsd.wait_for_attach() # Setup CUDA, GPU & distributed training if args.local_rank == -1 or args.no_cuda: device = torch.device("cuda" if torch.cuda.is_available() and not args.no_cuda else "cpu") args.n_gpu = torch.cuda.device_count() else: # Initializes the distributed backend which will take care of sychronizing nodes/GPUs torch.cuda.set_device(args.local_rank) device = torch.device("cuda", args.local_rank) torch.distributed.init_process_group(backend='nccl') args.n_gpu = 1 args.device = device args.per_gpu_train_batch_size=args.train_batch_size//args.n_gpu args.per_gpu_eval_batch_size=args.eval_batch_size//args.n_gpu # Setup logging logging.basicConfig(format='%(asctime)s - %(levelname)s - %(name)s - %(message)s', datefmt='%m/%d/%Y %H:%M:%S', level=logging.INFO if args.local_rank in [-1, 0] else logging.WARN) logger.warning("Process rank: %s, device: %s, n_gpu: %s, distributed training: %s, 16-bits training: %s", args.local_rank, device, args.n_gpu, bool(args.local_rank != -1), args.fp16) # Set seed set_seed(args.seed) # Load pretrained model and tokenizer if args.local_rank not in [-1, 0]: torch.distributed.barrier() # Barrier to make sure only the first process in distributed training download model & vocab args.start_epoch = 0 args.start_step = 0 checkpoint_last = os.path.join(args.output_dir, 'checkpoint-last') if os.path.exists(checkpoint_last) and os.listdir(checkpoint_last): args.model_name_or_path = os.path.join(checkpoint_last, 'pytorch_model.bin') args.config_name = os.path.join(checkpoint_last, 'config.json') idx_file = os.path.join(checkpoint_last, 'idx_file.txt') with open(idx_file, encoding='utf-8') as idxf: args.start_epoch = int(idxf.readlines()[0].strip()) + 1 step_file = os.path.join(checkpoint_last, 'step_file.txt') if os.path.exists(step_file): with open(step_file, encoding='utf-8') as stepf: args.start_step = int(stepf.readlines()[0].strip()) logger.info("reload model from {}, resume from {} epoch".format(checkpoint_last, args.start_epoch)) config_class, model_class, tokenizer_class = MODEL_CLASSES[args.model_type] config = config_class.from_pretrained(args.config_name if args.config_name else args.model_name_or_path, cache_dir=args.cache_dir if args.cache_dir else None) config.num_labels=1 tokenizer = tokenizer_class.from_pretrained(args.tokenizer_name, do_lower_case=args.do_lower_case, cache_dir=args.cache_dir if args.cache_dir else None) if args.block_size <= 0: args.block_size = tokenizer.max_len_single_sentence # Our input block size will be the max possible for the model args.block_size = min(args.block_size, tokenizer.max_len_single_sentence) if args.model_name_or_path: model = model_class.from_pretrained(args.model_name_or_path, from_tf=bool('.ckpt' in args.model_name_or_path), config=config, cache_dir=args.cache_dir if args.cache_dir else None) else: model = model_class(config) model=Model(model,config,tokenizer,args) if args.local_rank == 0: torch.distributed.barrier() # End of barrier to make sure only the first process in distributed training download model & vocab logger.info("Training/evaluation parameters %s", args) # Training if args.do_train: if args.local_rank not in [-1, 0]: torch.distributed.barrier() # Barrier to make sure only the first process in distributed training process the dataset, and the others will use the cache train_dataset = TextDataset(tokenizer, args,args.train_data_file) if args.local_rank == 0: torch.distributed.barrier() train(args, train_dataset, model, tokenizer) # Evaluation results = {} if args.do_eval and args.local_rank in [-1, 0]: checkpoint_prefix = 'epoch_5/subject_model.pth' output_dir = os.path.join(args.output_dir, '{}'.format(checkpoint_prefix)) model.load_state_dict(torch.load(output_dir)) model.to(args.device) result=evaluate(args, model, tokenizer) logger.info("***** Eval results *****") for key in sorted(result.keys()): logger.info(" %s = %s", key, str(round(result[key],4))) if args.do_test and args.local_rank in [-1, 0]: checkpoint_prefix = 'epoch_5/subject_model.pth' output_dir = os.path.join(args.output_dir, '{}'.format(checkpoint_prefix)) model.load_state_dict(torch.load(output_dir)) model.to(args.device) test(args, model, tokenizer) return results if __name__ == "__main__": main()