onnxruntime/orttraining/tools/scripts/nv_run_pretraining.py
Sherlock 6de79d82c8
Fix Training Packaging pipeline (#9885)
* Fix Training Packaging pipeline
2021-11-30 15:26:10 -08:00

635 lines
29 KiB
Python

# coding=utf-8
# Copyright (c) 2019 NVIDIA CORPORATION. All rights reserved.
# Copyright 2018 The Google AI Language Team Authors and The HugginFace Inc. team.
# 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.
"""BERT finetuning runner."""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
# ==================
import csv
import os
import time
import logging
import argparse
import random
import h5py
from tqdm import tqdm, trange
import os
import numpy as np
import torch
from torch.utils.data import DataLoader, RandomSampler, SequentialSampler, Dataset
from torch.utils.data.distributed import DistributedSampler
import math
from apex import amp
import multiprocessing
from tokenization import BertTokenizer
from modeling import BertForPreTraining, BertConfig
from optimization import BertLAMB
from file_utils import PYTORCH_PRETRAINED_BERT_CACHE
from utils import is_main_process
from apex.parallel import DistributedDataParallel as DDP
from schedulers import LinearWarmUpScheduler
from apex.parallel.distributed import flat_dist_call
import amp_C
import apex_C
from apex.amp import _amp_state
from concurrent.futures import ProcessPoolExecutor
logging.basicConfig(format='%(asctime)s - %(levelname)s - %(name)s - %(message)s',
datefmt='%m/%d/%Y %H:%M:%S',
level=logging.INFO)
logger = logging.getLogger(__name__)
def create_pretraining_dataset(input_file, max_pred_length, shared_list, args):
train_data = pretraining_dataset(input_file=input_file, max_pred_length=max_pred_length)
train_sampler = RandomSampler(train_data)
train_dataloader = DataLoader(train_data, sampler=train_sampler,
batch_size=args.train_batch_size * args.n_gpu, num_workers=4,
pin_memory=True)
# shared_list["0"] = (train_dataloader, input_file)
return train_dataloader, input_file
class pretraining_dataset(Dataset):
def __init__(self, input_file, max_pred_length):
self.input_file = input_file
self.max_pred_length = max_pred_length
f = h5py.File(input_file, "r")
keys = ['input_ids', 'input_mask', 'segment_ids', 'masked_lm_positions', 'masked_lm_ids',
'next_sentence_labels']
self.inputs = [np.asarray(f[key][:]) for key in keys]
f.close()
def __len__(self):
'Denotes the total number of samples'
return len(self.inputs[0])
def __getitem__(self, index):
[input_ids, input_mask, segment_ids, masked_lm_positions, masked_lm_ids, next_sentence_labels] = [
torch.from_numpy(input[index].astype(np.int64)) if indice < 5 else torch.from_numpy(
np.asarray(input[index].astype(np.int64))) for indice, input in enumerate(self.inputs)]
masked_lm_labels = torch.ones(input_ids.shape, dtype=torch.long) * -1
index = self.max_pred_length
# store number of masked tokens in index
padded_mask_indices = (masked_lm_positions == 0).nonzero()
if len(padded_mask_indices) != 0:
index = padded_mask_indices[0].item()
masked_lm_labels[masked_lm_positions[:index]] = masked_lm_ids[:index]
return [input_ids, segment_ids, input_mask,
masked_lm_labels, next_sentence_labels]
def parse_arguments():
parser = argparse.ArgumentParser()
## Required parameters
parser.add_argument("--input_dir",
default=None,
type=str,
required=True,
help="The input data dir. Should contain .hdf5 files for the task.")
parser.add_argument("--config_file",
default=None,
type=str,
required=True,
help="The BERT model config")
parser.add_argument("--bert_model", default="bert-large-uncased", type=str,
help="Bert pre-trained model selected in the list: bert-base-uncased, "
"bert-large-uncased, bert-base-cased, bert-base-multilingual, bert-base-chinese.")
parser.add_argument("--output_dir",
default=None,
type=str,
required=True,
help="The output directory where the model checkpoints will be written.")
## Other parameters
parser.add_argument("--max_seq_length",
default=512,
type=int,
help="The maximum total input sequence length after WordPiece tokenization. \n"
"Sequences longer than this will be truncated, and sequences shorter \n"
"than this will be padded.")
parser.add_argument("--max_predictions_per_seq",
default=80,
type=int,
help="The maximum total of masked tokens in input sequence")
parser.add_argument("--train_batch_size",
default=32,
type=int,
help="Total batch size for training.")
parser.add_argument("--learning_rate",
default=5e-5,
type=float,
help="The initial learning rate for Adam.")
parser.add_argument("--num_train_epochs",
default=3.0,
type=float,
help="Total number of training epochs to perform.")
parser.add_argument("--max_steps",
default=1000,
type=float,
help="Total number of training steps to perform.")
parser.add_argument("--warmup_proportion",
default=0.01,
type=float,
help="Proportion of training to perform linear learning rate warmup for. "
"E.g., 0.1 = 10%% of training.")
parser.add_argument("--local_rank",
type=int,
default=-1,
help="local_rank for distributed training on gpus")
parser.add_argument('--seed',
type=int,
default=42,
help="random seed for initialization")
parser.add_argument('--gradient_accumulation_steps',
type=int,
default=1,
help="Number of updates steps to accumualte before performing a backward/update pass.")
parser.add_argument('--fp16',
default=False,
action='store_true',
help="Whether to use 16-bit float precision instead of 32-bit")
parser.add_argument('--loss_scale',
type=float, default=0.0,
help='Loss scaling, positive power of 2 values can improve fp16 convergence.')
parser.add_argument('--log_freq',
type=float, default=50.0,
help='frequency of logging loss.')
parser.add_argument('--checkpoint_activations',
default=False,
action='store_true',
help="Whether to use gradient checkpointing")
parser.add_argument("--resume_from_checkpoint",
default=False,
action='store_true',
help="Whether to resume training from checkpoint.")
parser.add_argument('--resume_step',
type=int,
default=-1,
help="Step to resume training from.")
parser.add_argument('--num_steps_per_checkpoint',
type=int,
default=100,
help="Number of update steps until a model checkpoint is saved to disk.")
parser.add_argument('--phase2',
default=False,
action='store_true',
help="Whether to train with seq len 512")
parser.add_argument('--allreduce_post_accumulation',
default=False,
action='store_true',
help="Whether to do allreduces during gradient accumulation steps.")
parser.add_argument('--allreduce_post_accumulation_fp16',
default=False,
action='store_true',
help="Whether to do fp16 allreduce post accumulation.")
parser.add_argument('--accumulate_into_fp16',
default=False,
action='store_true',
help="Whether to use fp16 gradient accumulators.")
parser.add_argument('--phase1_end_step',
type=int,
default=7038,
help="Number of training steps in Phase1 - seq len 128")
parser.add_argument("--do_train",
default=False,
action='store_true',
help="Whether to run training.")
args = parser.parse_args()
return args
def setup_training(args):
assert (torch.cuda.is_available())
if args.local_rank == -1:
device = torch.device("cuda")
args.n_gpu = torch.cuda.device_count()
else:
torch.cuda.set_device(args.local_rank)
device = torch.device("cuda", args.local_rank)
args.n_gpu = 1
# Initializes the distributed backend which will take care of sychronizing nodes/GPUs
torch.distributed.init_process_group(backend='nccl', init_method='env://')
logger.info("device %s n_gpu %d distributed training %r", device, args.n_gpu, bool(args.local_rank != -1))
if args.gradient_accumulation_steps < 1:
raise ValueError("Invalid gradient_accumulation_steps parameter: {}, should be >= 1".format(
args.gradient_accumulation_steps))
if args.train_batch_size % args.gradient_accumulation_steps != 0:
raise ValueError("Invalid gradient_accumulation_steps parameter: {}, batch size {} should be divisible".format(
args.gradient_accumulation_steps, args.train_batch_size))
args.train_batch_size = args.train_batch_size // args.gradient_accumulation_steps
if not args.do_train:
raise ValueError(" `do_train` must be True.")
if not args.resume_from_checkpoint and os.path.exists(args.output_dir) and (
os.listdir(args.output_dir) and os.listdir(args.output_dir) != ['logfile.txt']):
raise ValueError("Output directory ({}) already exists and is not empty.".format(args.output_dir))
if not args.resume_from_checkpoint:
os.makedirs(args.output_dir, exist_ok=True)
return device, args
def prepare_model_and_optimizer(args, device):
# Prepare model
config = BertConfig.from_json_file(args.config_file)
# Padding for divisibility by 8
if config.vocab_size % 8 != 0:
config.vocab_size += 8 - (config.vocab_size % 8)
model = BertForPreTraining(config)
checkpoint = None
if not args.resume_from_checkpoint:
global_step = 0
else:
if args.resume_step == -1:
model_names = [f for f in os.listdir(args.output_dir) if f.endswith(".pt")]
args.resume_step = max([int(x.split('.pt')[0].split('_')[1].strip()) for x in model_names])
global_step = args.resume_step
checkpoint = torch.load(os.path.join(args.output_dir, "ckpt_{}.pt".format(global_step)), map_location="cpu")
model.load_state_dict(checkpoint['model'], strict=False)
if args.phase2:
global_step -= args.phase1_end_step
if is_main_process():
print("resume step from ", args.resume_step)
model.to(device)
param_optimizer = list(model.named_parameters())
no_decay = ['bias', 'gamma', 'beta', 'LayerNorm']
optimizer_grouped_parameters = []
names = []
count = 1
for n, p in param_optimizer:
count += 1
if not any(nd in n for nd in no_decay):
optimizer_grouped_parameters.append({'params': [p], 'weight_decay': 0.01, 'name': n})
names.append({'params': [n], 'weight_decay': 0.01})
if any(nd in n for nd in no_decay):
optimizer_grouped_parameters.append({'params': [p], 'weight_decay': 0.00, 'name': n})
names.append({'params': [n], 'weight_decay': 0.00})
optimizer = BertLAMB(optimizer_grouped_parameters,
lr=args.learning_rate,
warmup=args.warmup_proportion,
t_total=args.max_steps)
if args.fp16:
if args.loss_scale == 0:
# optimizer = FP16_Optimizer(optimizer, dynamic_loss_scale=True)
model, optimizer = amp.initialize(model, optimizer, opt_level="O2", loss_scale="dynamic",
master_weights=False if args.accumulate_into_fp16 else True)
else:
# optimizer = FP16_Optimizer(optimizer, static_loss_scale=args.loss_scale)
model, optimizer = amp.initialize(model, optimizer, opt_level="O2", loss_scale=args.loss_scale,
master_weights=False if args.accumulate_into_fp16 else True)
amp._amp_state.loss_scalers[0]._loss_scale = 2**20
if args.resume_from_checkpoint:
if args.phase2:
keys = list(checkpoint['optimizer']['state'].keys())
#Override hyperparameters from Phase 1
for key in keys:
checkpoint['optimizer']['state'][key]['step'] = global_step
for iter, item in enumerate(checkpoint['optimizer']['param_groups']):
checkpoint['optimizer']['param_groups'][iter]['t_total'] = args.max_steps
checkpoint['optimizer']['param_groups'][iter]['warmup'] = args.warmup_proportion
checkpoint['optimizer']['param_groups'][iter]['lr'] = args.learning_rate
optimizer.load_state_dict(checkpoint['optimizer']) # , strict=False)
# Restore AMP master parameters
if args.fp16:
optimizer._lazy_init_maybe_master_weights()
optimizer._amp_stash.lazy_init_called = True
optimizer.load_state_dict(checkpoint['optimizer'])
for param, saved_param in zip(amp.master_params(optimizer), checkpoint['master params']):
param.data.copy_(saved_param.data)
if args.local_rank != -1:
if not args.allreduce_post_accumulation:
model = DDP(model, message_size=250000000, gradient_predivide_factor=torch.distributed.get_world_size())
else:
flat_dist_call([param.data for param in model.parameters()], torch.distributed.broadcast, (0,) )
elif args.n_gpu > 1:
model = torch.nn.DataParallel(model)
return model, optimizer, checkpoint, global_step
def take_optimizer_step(args, optimizer, model, overflow_buf, global_step):
if args.allreduce_post_accumulation:
# manually allreduce gradients after all accumulation steps
# check for Inf/NaN
# 1. allocate an uninitialized buffer for flattened gradient
scaler = _amp_state.loss_scalers[0]
master_grads = [p.grad for p in amp.master_params(optimizer) if p.grad is not None]
flat_grad_size = sum(p.numel() for p in master_grads)
allreduce_dtype = torch.float16 if args.allreduce_post_accumulation_fp16 else torch.float32
flat_raw = torch.empty(flat_grad_size, device='cuda', dtype=allreduce_dtype)
# 2. combine unflattening and predivision of unscaled 'raw' gradient
allreduced_views = apex_C.unflatten(flat_raw, master_grads)
overflow_buf.zero_()
amp_C.multi_tensor_scale(65536,
overflow_buf,
[master_grads, allreduced_views],
scaler.loss_scale() / (torch.distributed.get_world_size() * args.gradient_accumulation_steps))
# 3. sum gradient across ranks. Because of the predivision, this averages the gradient
torch.distributed.all_reduce(flat_raw)
# 4. combine unscaling and unflattening of allreduced gradient
overflow_buf.zero_()
amp_C.multi_tensor_scale(65536,
overflow_buf,
[allreduced_views, master_grads],
1./scaler.loss_scale())
# 5. update loss scale
scaler = _amp_state.loss_scalers[0]
old_overflow_buf = scaler._overflow_buf
scaler._overflow_buf = overflow_buf
had_overflow = scaler.update_scale()
scaler._overfloat_buf = old_overflow_buf
# 6. call optimizer step function
if had_overflow == 0:
optimizer.step()
global_step += 1
else:
# Overflow detected, print message and clear gradients
if is_main_process():
print(("Rank {} :: Gradient overflow. Skipping step, " +
"reducing loss scale to {}").format(
torch.distributed.get_rank(),
scaler.loss_scale()))
if _amp_state.opt_properties.master_weights:
for param in optimizer._amp_stash.all_fp32_from_fp16_params:
param.grad = None
for param in model.parameters():
param.grad = None
else:
optimizer.step()
#optimizer.zero_grad()
for param in model.parameters():
param.grad = None
global_step += 1
return global_step
def main():
args = parse_arguments()
random.seed(args.seed)
np.random.seed(args.seed)
torch.manual_seed(args.seed)
device, args = setup_training(args)
# Prepare optimizer
config = BertConfig.from_json_file(args.config_file)
model, optimizer, checkpoint, global_step = prepare_model_and_optimizer(args, device)
is_model_exported = False
if is_main_process():
print("SEED {}".format(args.seed))
if args.do_train:
if is_main_process():
logger.info("***** Running training *****")
# logger.info(" Num examples = %d", len(train_data))
logger.info(" Batch size = %d", args.train_batch_size)
print(" LR = ", args.learning_rate)
print("Training. . .")
model.train()
most_recent_ckpts_paths = []
average_loss = 0.0 # averaged loss every args.log_freq steps
epoch = 0
training_steps = 0
pool = ProcessPoolExecutor(1)
# Note: We loop infinitely over epochs, termination is handled via iteration count
while True:
thread = None
if not args.resume_from_checkpoint or epoch > 0 or args.phase2:
files = [os.path.join(args.input_dir, f) for f in os.listdir(args.input_dir) if
os.path.isfile(os.path.join(args.input_dir, f))]
files.sort()
num_files = len(files)
random.shuffle(files)
f_start_id = 0
else:
f_start_id = checkpoint['files'][0]
files = checkpoint['files'][1:]
args.resume_from_checkpoint = False
num_files = len(files)
print('File list is [' + ','.join(files) + '].')
shared_file_list = {}
if torch.distributed.is_initialized() and torch.distributed.get_world_size() > num_files:
remainder = torch.distributed.get_world_size() % num_files
data_file = files[(f_start_id*torch.distributed.get_world_size()+torch.distributed.get_rank() + remainder*f_start_id)%num_files]
else:
data_file = files[f_start_id%num_files]
previous_file = data_file
print('Create pretraining_dataset with file {}...'.format(data_file))
train_data = pretraining_dataset(data_file, args.max_predictions_per_seq)
train_sampler = RandomSampler(train_data)
train_dataloader = DataLoader(train_data, sampler=train_sampler,
batch_size=args.train_batch_size * args.n_gpu, num_workers=4,
pin_memory=True)
# shared_file_list["0"] = (train_dataloader, data_file)
overflow_buf = None
if args.allreduce_post_accumulation:
overflow_buf = torch.cuda.IntTensor([0])
for f_id in range(f_start_id + 1 , len(files)):
# torch.cuda.synchronize()
# f_start = time.time()
if torch.distributed.is_initialized() and torch.distributed.get_world_size() > num_files:
data_file = files[(f_id*torch.distributed.get_world_size()+torch.distributed.get_rank() + remainder*f_id)%num_files]
else:
data_file = files[f_id%num_files]
logger.info("file no %s file %s" % (f_id, previous_file))
previous_file = data_file
# train_dataloader = shared_file_list["0"][0]
# thread = multiprocessing.Process(
# name="LOAD DATA:" + str(f_id) + ":" + str(data_file),
# target=create_pretraining_dataset,
# args=(data_file, args.max_predictions_per_seq, shared_file_list, args, n_gpu)
# )
# thread.start()
print('Submit new data file {0} for the next iteration...'.format(data_file))
dataset_future = pool.submit(create_pretraining_dataset, data_file, args.max_predictions_per_seq, shared_file_list, args)
# torch.cuda.synchronize()
# f_end = time.time()
# print('[{}] : shard overhead {}'.format(torch.distributed.get_rank(), f_end - f_start))
train_iter = tqdm(train_dataloader, desc="Iteration") if is_main_process() else train_dataloader
for step, batch in enumerate(train_iter):
# torch.cuda.synchronize()
# iter_start = time.time()
training_steps += 1
batch = [t.to(device) for t in batch]
input_ids, segment_ids, input_mask, masked_lm_labels, next_sentence_labels = batch
if not is_model_exported:
onnx_path = os.path.join(args.output_dir, 'bert_for_pretraining_without_loss_' + config.to_string() + '.onnx')
lm_score, sq_score = model(input_ids=input_ids, token_type_ids=segment_ids, attention_mask=input_mask)
torch.onnx.export(model, (input_ids, segment_ids, input_mask), onnx_path,
verbose = True,
#input_names = ['input_ids', 'token_type_ids', 'input_mask'],
input_names = ['input1', 'input2', 'input3'],
output_names = ['output1', 'output2'],
dynamic_axes={'input1': {0: 'batch'}, 'input2': {0: 'batch'}, 'input3': {0: 'batch'}, 'output1': {0: 'batch'}, 'output2': {0: 'batch'}},
training=True)
is_model_exported = False
import onnxruntime as ort
sess = ort.InferenceSession(onnx_path, providers=ort.get_available_providers())
result = sess.run(None, {'input1': input_ids.cpu().numpy(), 'input2': segment_ids.cpu().numpy(), 'input3': input_mask.cpu().numpy()})
print('---ORT result---')
print(result[0])
print(result[1])
print('---Pytorch result---')
print(lm_score)
print(sq_score)
print('---ORT-Pytorch Diff---')
print(np.linalg.norm(result[0]-lm_score.detach().cpu().numpy()))
print(np.linalg.norm(result[1]-sq_score.detach().cpu().numpy()))
return
loss = model(input_ids=input_ids, token_type_ids=segment_ids, attention_mask=input_mask,
masked_lm_labels=masked_lm_labels, next_sentence_label=next_sentence_labels,
checkpoint_activations=args.checkpoint_activations)
if args.n_gpu > 1:
loss = loss.mean() # mean() to average on multi-gpu.
divisor = args.gradient_accumulation_steps
if args.gradient_accumulation_steps > 1:
if not args.allreduce_post_accumulation:
# this division was merged into predivision
loss = loss / args.gradient_accumulation_steps
divisor = 1.0
if args.fp16:
with amp.scale_loss(loss, optimizer, delay_overflow_check=args.allreduce_post_accumulation) as scaled_loss:
scaled_loss.backward()
else:
loss.backward()
average_loss += loss.item()
if training_steps % args.gradient_accumulation_steps == 0:
global_step = take_optimizer_step(args, optimizer, model, overflow_buf, global_step)
if global_step >= args.max_steps:
last_num_steps = global_step % args.log_freq
last_num_steps = args.log_freq if last_num_steps == 0 else last_num_steps
average_loss = torch.tensor(average_loss, dtype=torch.float32).cuda()
average_loss = average_loss / (last_num_steps * divisor)
if (torch.distributed.is_initialized()):
average_loss /= torch.distributed.get_world_size()
torch.distributed.all_reduce(average_loss)
if is_main_process():
logger.info("Total Steps:{} Final Loss = {}".format(training_steps, average_loss.item()))
elif training_steps % (args.log_freq * args.gradient_accumulation_steps) == 0:
if is_main_process():
print("Step:{} Average Loss = {} Step Loss = {} LR {}".format(global_step, average_loss / (
args.log_freq * divisor),
loss.item() * args.gradient_accumulation_steps / divisor,
optimizer.param_groups[0][
'lr']))
average_loss = 0
if global_step >= args.max_steps or training_steps % (
args.num_steps_per_checkpoint * args.gradient_accumulation_steps) == 0:
if is_main_process():
# Save a trained model
logger.info("** ** * Saving fine - tuned model ** ** * ")
model_to_save = model.module if hasattr(model,
'module') else model # Only save the model it-self
if args.resume_step < 0 or not args.phase2:
output_save_file = os.path.join(args.output_dir, "ckpt_{}.pt".format(global_step))
else:
output_save_file = os.path.join(args.output_dir, "ckpt_{}.pt".format(global_step + args.phase1_end_step))
if args.do_train:
torch.save({'model': model_to_save.state_dict(),
'optimizer': optimizer.state_dict(),
'master params': list(amp.master_params(optimizer)),
'files': [f_id] + files}, output_save_file)
most_recent_ckpts_paths.append(output_save_file)
if len(most_recent_ckpts_paths) > 3:
ckpt_to_be_removed = most_recent_ckpts_paths.pop(0)
os.remove(ckpt_to_be_removed)
if global_step >= args.max_steps:
del train_dataloader
# thread.join()
return args
# torch.cuda.synchronize()
# iter_end = time.time()
# if torch.distributed.get_rank() == 0:
# print('step {} : {}'.format(global_step, iter_end - iter_start))
del train_dataloader
# thread.join()
# Make sure pool has finished and switch train_dataloader
# NOTE: Will block until complete
train_dataloader, data_file = dataset_future.result(timeout=None)
epoch += 1
if __name__ == "__main__":
now = time.time()
args = main()
if is_main_process():
print("Total time taken {}".format(time.time() - now))