xtuner
xtuner copied to clipboard
Published
20 hours ago •
InternLM
InternLM
请问是否支持200k上文的微调，需要什么样的配置？
目前我使用的是zero3对interlm2-chat-20b进行全量微调，8*A100只能微调2k上文，请问200k或者比较长几十k的上文微调需要怎么设置？
目前 XTuner 支持使用 8卡 80G A100 训练 8k上下文的 20b 模型（在 config 中设置 max_length = 8192 即可）。训练 32k 上下文的 20b 模型需要 32卡 zero3。 我们正在开发 200k 上下文训练 20b 模型的解决方案，近期会发布。
目前 XTuner 支持使用 8卡 80G A100 训练 8k上下文的 20b 模型（在 config 中设置 max_length = 8192 即可）。训练 32k 上下文的 20b 模型需要 32卡 zero3。 我们正在开发 200k 上下文训练 20b 模型的解决方案，近期会发布。
您好，目前我是根据项目给出的XTuner脚本改动为全量FT后进行微调，目前2k的上文需要8*80GA100加上ZeRO3进行微调，并且显存基本已经用满。我在internlm项目issue中提问得到的解答是100k上文可以通过设置tensor parallelism，pipeline parallelism，gradient checkpointing这些参数进行8卡微调，但是我没有找到如何设置。 根据您的解答我如果想要微调8k上文直接修改我XTuner的训练配置中max_length = 8192即可，是否需要其它更改？ 我在微调2k上文时使用的命令是： NPROC_PER_NODE=8 xtuner train internlm2_chat_20b_full_alpaca_e3.py --deepspeed deepspeed_zero3 以下是我微调2k上文时的脚本：
# Copyright (c) OpenMMLab. All rights reserved.
import torch
from datasets import load_dataset
from mmengine.dataset import DefaultSampler
from mmengine.hooks import (CheckpointHook, DistSamplerSeedHook, IterTimerHook,
LoggerHook, ParamSchedulerHook)
from mmengine.optim import AmpOptimWrapper, CosineAnnealingLR, LinearLR
from peft import LoraConfig
from torch.optim import AdamW
from transformers import (AutoModelForCausalLM, AutoTokenizer,
BitsAndBytesConfig)
from xtuner.dataset import process_hf_dataset
from xtuner.dataset.collate_fns import default_collate_fn
from xtuner.dataset.map_fns import alpaca_map_fn, template_map_fn_factory
from xtuner.engine import DatasetInfoHook, EvaluateChatHook
from xtuner.model import SupervisedFinetune
from xtuner.utils import PROMPT_TEMPLATE, SYSTEM_TEMPLATE
#######################################################################
# PART 1 Settings
#######################################################################
# Model
pretrained_model_name_or_path = './internlm2-chat-20b'
# Data
alpaca_en_path = './myalpaca'
prompt_template = PROMPT_TEMPLATE.internlm2_chat
max_length = 2048
pack_to_max_length = True
# Scheduler & Optimizer
batch_size = 1 # per_device
accumulative_counts = 1
dataloader_num_workers = 0
max_epochs = 3
optim_type = AdamW
lr = 2e-4
betas = (0.9, 0.999)
weight_decay = 0
max_norm = 1 # grad clip
warmup_ratio = 0.03
# Evaluate the generation performance during the training
evaluation_freq = 500
SYSTEM = SYSTEM_TEMPLATE.alpaca
evaluation_inputs = [
'请给我介绍五个上海的景点', 'Please tell me five scenic spots in Shanghai'
]
#######################################################################
# PART 2 Model & Tokenizer #
#######################################################################
tokenizer = dict(
type=AutoTokenizer.from_pretrained,
pretrained_model_name_or_path=pretrained_model_name_or_path,
trust_remote_code=True,
padding_side='right')
model = dict(
type=SupervisedFinetune,
llm=dict(
type=AutoModelForCausalLM.from_pretrained,
pretrained_model_name_or_path=pretrained_model_name_or_path,
trust_remote_code=True,
torch_dtype=torch.float16,
# quantization_config=dict(
# type=BitsAndBytesConfig,
# load_in_4bit=True,
# load_in_8bit=False,
# llm_int8_threshold=6.0,
# llm_int8_has_fp16_weight=False,
# bnb_4bit_compute_dtype=torch.float16,
# bnb_4bit_use_double_quant=True,
# bnb_4bit_quant_type='nf4')),
# lora=dict(
# type=LoraConfig,
# r=64,
# lora_alpha=16,
# lora_dropout=0.1,
# bias='none',
# task_type='CAUSAL_LM')
))
#######################################################################
# PART 3 Dataset & Dataloader #
#######################################################################
alpaca_en = dict(
type=process_hf_dataset,
dataset=dict(type=load_dataset, path=alpaca_en_path),
tokenizer=tokenizer,
max_length=max_length,
dataset_map_fn=alpaca_map_fn,
template_map_fn=dict(
type=template_map_fn_factory, template=prompt_template),
remove_unused_columns=True,
shuffle_before_pack=True,
pack_to_max_length=pack_to_max_length)
train_dataloader = dict(
batch_size=batch_size,
num_workers=dataloader_num_workers,
dataset=alpaca_en,
sampler=dict(type=DefaultSampler, shuffle=True),
collate_fn=dict(type=default_collate_fn))
#######################################################################
# PART 4 Scheduler & Optimizer #
#######################################################################
# optimizer
optim_wrapper = dict(
type=AmpOptimWrapper,
optimizer=dict(
type=optim_type, lr=lr, betas=betas, weight_decay=weight_decay),
clip_grad=dict(max_norm=max_norm, error_if_nonfinite=False),
accumulative_counts=accumulative_counts,
loss_scale='dynamic',
dtype='float16')
# learning policy
# More information: https://github.com/open-mmlab/mmengine/blob/main/docs/en/tutorials/param_scheduler.md # noqa: E501
param_scheduler = [
dict(
type=LinearLR,
start_factor=1e-5,
by_epoch=True,
begin=0,
end=warmup_ratio * max_epochs,
convert_to_iter_based=True),
dict(
type=CosineAnnealingLR,
eta_min=0.0,
by_epoch=True,
begin=warmup_ratio * max_epochs,
T_max=max_epochs,
convert_to_iter_based=True)
]
# train, val, test setting
train_cfg = dict(by_epoch=True, max_epochs=max_epochs, val_interval=1)
#######################################################################
# PART 5 Runtime #
#######################################################################
# Log the dialogue periodically during the training process, optional
custom_hooks = [
dict(type=DatasetInfoHook, tokenizer=tokenizer),
dict(
type=EvaluateChatHook,
tokenizer=tokenizer,
every_n_iters=evaluation_freq,
evaluation_inputs=evaluation_inputs,
system=SYSTEM,
prompt_template=prompt_template)
]
# configure default hooks
default_hooks = dict(
# record the time of every iteration.
timer=dict(type=IterTimerHook),
# print log every 100 iterations.
logger=dict(type=LoggerHook, interval=10),
# enable the parameter scheduler.
param_scheduler=dict(type=ParamSchedulerHook),
# save checkpoint per epoch.
checkpoint=dict(type=CheckpointHook, interval=1),
# set sampler seed in distributed evrionment.
sampler_seed=dict(type=DistSamplerSeedHook),
)
# configure environment
env_cfg = dict(
# whether to enable cudnn benchmark
cudnn_benchmark=False,
# set multi process parameters
mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0),
# set distributed parameters
dist_cfg=dict(backend='nccl'),
)
# set visualizer
visualizer = None
# set log level
log_level = 'INFO'
# load from which checkpoint
load_from = None
# whether to resume training from the loaded checkpoint
resume = False
# Defaults to use random seed and disable `deterministic`
randomness = dict(seed=None, deterministic=False)
非常感谢🙏，期待您的解答
XTuner 训练的时候默认是会打开 gradient checkpointing 的。tensor parallelism，pipeline parallelism目前 XTuner 还没支持，但我们正在支持 Sequence Parallelism 以实现超长上下文的训练，近期会发布。
使用8卡跑Internlm2 20B模型的微调，推荐基于internlm_7b_full_alpaca_e3.py进行修改。
- pretrained_model_name_or_path = 'internlm/internlm-7b'
- use_varlen_attn = False
+ pretrained_model_name_or_path = 'internlm/internlm2-20b'
+ use_varlen_attn = True
- max_length = 2048
+ max_length = 8192
我这里有一份我的开发机可以跑的config供你参考：
# Copyright (c) OpenMMLab. All rights reserved.
from datasets import load_dataset
from mmengine.dataset import DefaultSampler
from mmengine.hooks import (CheckpointHook, DistSamplerSeedHook, IterTimerHook,
LoggerHook, ParamSchedulerHook)
from mmengine.optim import AmpOptimWrapper, CosineAnnealingLR, LinearLR
from torch.optim import AdamW
from transformers import AutoModelForCausalLM, AutoTokenizer
from xtuner.dataset import process_hf_dataset
from xtuner.dataset.collate_fns import default_collate_fn
from xtuner.dataset.map_fns import alpaca_map_fn, template_map_fn_factory
from xtuner.engine.hooks import (DatasetInfoHook, EvaluateChatHook,
ThroughputHook,
VarlenAttnArgsToMessageHubHook)
from xtuner.engine.runner import TrainLoop
from xtuner.model import SupervisedFinetune
from xtuner.utils import PROMPT_TEMPLATE, SYSTEM_TEMPLATE
#######################################################################
# PART 1 Settings #
#######################################################################
# Model
pretrained_model_name_or_path = 'internlm/internlm2-20b' # noqa: E501
use_varlen_attn = True
# Data
alpaca_en_path = 'tatsu-lab/alpaca'
prompt_template = PROMPT_TEMPLATE.internlm2_chat
max_length = 8192
pack_to_max_length = True
# Scheduler & Optimizer
batch_size = 1 # per_device
accumulative_counts = 1
dataloader_num_workers = 0
max_epochs = 3
optim_type = AdamW
lr = 4e-5
betas = (0.9, 0.95)
weight_decay = 0.01
max_norm = 1 # grad clip
warmup_ratio = 0.025
# Save
save_steps = 100
save_total_limit = 2 # Maximum checkpoints to keep (-1 means unlimited)
# Evaluate the generation performance during the training
evaluation_freq = 50
SYSTEM = SYSTEM_TEMPLATE.alpaca
evaluation_inputs = [
'请给我介绍五个上海的景点', 'Please tell me five scenic spots in Shanghai'
]
#######################################################################
# PART 2 Model & Tokenizer #
#######################################################################
tokenizer = dict(
type=AutoTokenizer.from_pretrained,
pretrained_model_name_or_path=pretrained_model_name_or_path,
trust_remote_code=True,
padding_side='right')
model = dict(
type=SupervisedFinetune,
use_varlen_attn=use_varlen_attn,
llm=dict(
type=AutoModelForCausalLM.from_pretrained,
pretrained_model_name_or_path=pretrained_model_name_or_path,
trust_remote_code=True))
#######################################################################
# PART 3 Dataset & Dataloader #
#######################################################################
alpaca_en = dict(
type=process_hf_dataset,
dataset=dict(type=load_dataset, path=alpaca_en_path),
tokenizer=tokenizer,
max_length=max_length,
dataset_map_fn=alpaca_map_fn,
template_map_fn=dict(
type=template_map_fn_factory, template=prompt_template),
remove_unused_columns=True,
shuffle_before_pack=True,
pack_to_max_length=pack_to_max_length,
use_varlen_attn=use_varlen_attn)
train_dataloader = dict(
batch_size=batch_size,
num_workers=dataloader_num_workers,
dataset=alpaca_en,
sampler=dict(type=DefaultSampler, shuffle=True),
collate_fn=dict(type=default_collate_fn, use_varlen_attn=use_varlen_attn))
#######################################################################
# PART 4 Scheduler & Optimizer #
#######################################################################
# optimizer
optim_wrapper = dict(
type=AmpOptimWrapper,
optimizer=dict(
type=optim_type, lr=lr, betas=betas, weight_decay=weight_decay),
clip_grad=dict(max_norm=max_norm, error_if_nonfinite=False),
accumulative_counts=accumulative_counts,
loss_scale='dynamic',
dtype='float16')
# learning policy
# More information: https://github.com/open-mmlab/mmengine/blob/main/docs/en/tutorials/param_scheduler.md # noqa: E501
param_scheduler = [
dict(
type=LinearLR,
start_factor=1e-5,
by_epoch=True,
begin=0,
end=warmup_ratio * max_epochs,
convert_to_iter_based=True),
dict(
type=CosineAnnealingLR,
eta_min=0.0,
by_epoch=True,
begin=warmup_ratio * max_epochs,
end=max_epochs,
convert_to_iter_based=True)
]
# train, val, test setting
train_cfg = dict(type=TrainLoop, max_epochs=max_epochs)
#######################################################################
# PART 5 Runtime #
#######################################################################
# Log the dialogue periodically during the training process, optional
custom_hooks = [
dict(type=DatasetInfoHook, tokenizer=tokenizer),
dict(
type=EvaluateChatHook,
tokenizer=tokenizer,
every_n_iters=evaluation_freq,
evaluation_inputs=evaluation_inputs,
system=SYSTEM,
prompt_template=prompt_template),
dict(type=ThroughputHook)
]
if use_varlen_attn:
custom_hooks += [dict(type=VarlenAttnArgsToMessageHubHook)]
# configure default hooks
default_hooks = dict(
# record the time of every iteration.
timer=dict(type=IterTimerHook),
# print log every 10 iterations.
logger=dict(type=LoggerHook, log_metric_by_epoch=False, interval=1),
# enable the parameter scheduler.
param_scheduler=dict(type=ParamSchedulerHook),
# save checkpoint per `save_steps`.
checkpoint=dict(
type=CheckpointHook,
by_epoch=False,
interval=save_steps,
max_keep_ckpts=save_total_limit),
# set sampler seed in distributed evrionment.
sampler_seed=dict(type=DistSamplerSeedHook),
)
# configure environment
env_cfg = dict(
# whether to enable cudnn benchmark
cudnn_benchmark=False,
# set multi process parameters
mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0),
# set distributed parameters
dist_cfg=dict(backend='nccl'),
)
# set visualizer
visualizer = None
# set log level
log_level = 'INFO'
# load from which checkpoint
load_from = None
# whether to resume training from the loaded checkpoint
resume = False
# Defaults to use random seed and disable `deterministic`
randomness = dict(seed=None, deterministic=False)
# set log processor
log_processor = dict(by_epoch=False)