github
/
ColossalAI
mirror of https://github.com/hpcaitech/ColossalAI
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

update lightning version (#1954)

pull/1958/head
Fazzie-Maqianli 2022-11-15 16:57:48 +08:00 committed by GitHub
parent 52c6ad26e0
commit 6bdd0a90ca
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
13 changed files with 29 additions and 35 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
examples/images/diffusion/README.md
View File

@ -44,13 +44,6 @@ pip install -e .
pip install colossalai==0.1.10+torch1.11cu11.3 -f https://release.colossalai.org pip install colossalai==0.1.10+torch1.11cu11.3 -f https://release.colossalai.org
``` ```
### Install [Lightning](https://github.com/Lightning-AI/lightning)
We use the Sep. 2022 version with commit id as `b04a7aa`.
```
git clone https://github.com/Lightning-AI/lightning && cd lightning && git reset --hard b04a7aa
pip install -r requirements.txt && pip install .
```
> The specified version is due to the interface incompatibility caused by the latest update of [Lightning](https://github.com/Lightning-AI/lightning), which will be fixed in the near future. > The specified version is due to the interface incompatibility caused by the latest update of [Lightning](https://github.com/Lightning-AI/lightning), which will be fixed in the near future.
## Dataset ## Dataset

4
examples/images/diffusion/configs/train_colossalai.yaml
View File

@ -94,7 +94,7 @@ lightning:
precision: 16 precision: 16
auto_select_gpus: False auto_select_gpus: False
strategy: strategy:
target: pytorch_lightning.strategies.ColossalAIStrategy target: lightning.pytorch.strategies.ColossalAIStrategy
params: params:
use_chunk: False use_chunk: False
enable_distributed_storage: True, enable_distributed_storage: True,
@ -108,7 +108,7 @@ lightning:
logger_config: logger_config:
wandb: wandb:
target: pytorch_lightning.loggers.WandbLogger target: lightning.pytorch.loggers.WandbLogger
params: params:
name: nowname name: nowname
save_dir: "/tmp/diff_log/" save_dir: "/tmp/diff_log/"

4
examples/images/diffusion/configs/train_colossalai_cifar10.yaml
View File

@ -101,7 +101,7 @@ lightning:
precision: 16 precision: 16
auto_select_gpus: False auto_select_gpus: False
strategy: strategy:
target: pytorch_lightning.strategies.ColossalAIStrategy target: lightning.pytorch.strategies.ColossalAIStrategy
params: params:
use_chunk: False use_chunk: False
enable_distributed_storage: True, enable_distributed_storage: True,
@ -115,7 +115,7 @@ lightning:
logger_config: logger_config:
wandb: wandb:
target: pytorch_lightning.loggers.WandbLogger target: lightning.pytorch.loggers.WandbLogger
params: params:
name: nowname name: nowname
save_dir: "/tmp/diff_log/" save_dir: "/tmp/diff_log/"

4
examples/images/diffusion/configs/train_ddp.yaml
View File

@ -94,7 +94,7 @@ lightning:
precision: 16 precision: 16
auto_select_gpus: False auto_select_gpus: False
strategy: strategy:
target: pytorch_lightning.strategies.DDPStrategy target: lightning.pytorch.strategies.DDPStrategy
params: params:
find_unused_parameters: False find_unused_parameters: False
log_every_n_steps: 2 log_every_n_steps: 2
@ -105,7 +105,7 @@ lightning:
logger_config: logger_config:
wandb: wandb:
target: pytorch_lightning.loggers.WandbLogger target: lightning.pytorch.loggers.WandbLogger
params: params:
name: nowname name: nowname
save_dir: "/tmp/diff_log/" save_dir: "/tmp/diff_log/"

4
examples/images/diffusion/configs/train_pokemon.yaml
View File

@ -95,7 +95,7 @@ lightning:
precision: 16 precision: 16
auto_select_gpus: False auto_select_gpus: False
strategy: strategy:
target: pytorch_lightning.strategies.ColossalAIStrategy target: lightning.pytorch.strategies.ColossalAIStrategy
params: params:
use_chunk: False use_chunk: False
enable_distributed_storage: True, enable_distributed_storage: True,
@ -113,7 +113,7 @@ lightning:
logger_config: logger_config:
wandb: wandb:
target: pytorch_lightning.loggers.WandbLogger target: lightning.pytorch.loggers.WandbLogger
params: params:
name: nowname name: nowname
save_dir: "/tmp/diff_log/" save_dir: "/tmp/diff_log/"

2
examples/images/diffusion/environment.yaml
View File

@ -18,7 +18,7 @@ dependencies:
- invisible-watermark - invisible-watermark
- imageio==2.9.0 - imageio==2.9.0
- imageio-ffmpeg==0.4.2 - imageio-ffmpeg==0.4.2
- pytorch-lightning==1.8.0 - lightning==1.8.1
- omegaconf==2.1.1 - omegaconf==2.1.1
- test-tube>=0.7.5 - test-tube>=0.7.5
- streamlit>=0.73.1 - streamlit>=0.73.1

2
examples/images/diffusion/ldm/models/autoencoder.py
View File

@ -1,5 +1,5 @@
import torch import torch
import pytorch_lightning as pl import lightning.pytorch as pl
import torch.nn.functional as F import torch.nn.functional as F
from contextlib import contextmanager from contextlib import contextmanager

2
examples/images/diffusion/ldm/models/diffusion/classifier.py
View File

@ -1,6 +1,6 @@
import os import os
import torch import torch
import pytorch_lightning as pl import lightning.pytorch as pl
from omegaconf import OmegaConf from omegaconf import OmegaConf
from torch.nn import functional as F from torch.nn import functional as F
from torch.optim import AdamW from torch.optim import AdamW

6
examples/images/diffusion/ldm/models/diffusion/ddpm.py
View File

@ -1,7 +1,7 @@
import torch import torch
import torch.nn as nn import torch.nn as nn
import numpy as np import numpy as np
import pytorch_lightning as pl import lightning.pytorch as pl
from torch.optim.lr_scheduler import LambdaLR from torch.optim.lr_scheduler import LambdaLR
from einops import rearrange, repeat from einops import rearrange, repeat
from contextlib import contextmanager from contextlib import contextmanager
@ -9,8 +9,8 @@ from functools import partial
from tqdm import tqdm from tqdm import tqdm
from torchvision.utils import make_grid from torchvision.utils import make_grid
from pytorch_lightning.utilities.rank_zero import rank_zero_only from lightning.pytorch.utilities.rank_zero import rank_zero_only
from pytorch_lightning.utilities import rank_zero_info from lightning.pytorch.utilities import rank_zero_info
from ldm.util import log_txt_as_img, exists, default, ismap, isimage, mean_flat, count_params, instantiate_from_config from ldm.util import log_txt_as_img, exists, default, ismap, isimage, mean_flat, count_params, instantiate_from_config
from ldm.modules.ema import LitEma from ldm.modules.ema import LitEma

24
examples/images/diffusion/main.py
View File

@ -3,23 +3,23 @@ import numpy as np
import time import time
import torch import torch
import torchvision import torchvision
import pytorch_lightning as pl import lightning.pytorch as pl
from packaging import version from packaging import version
from omegaconf import OmegaConf from omegaconf import OmegaConf
from torch.utils.data import random_split, DataLoader, Dataset, Subset from torch.utils.data import random_split, DataLoader, Dataset, Subset
from functools import partial from functools import partial
from PIL import Image from PIL import Image
# from pytorch_lightning.strategies.colossalai import ColossalAIStrategy # from lightning.pytorch.strategies.colossalai import ColossalAIStrategy
# from colossalai.nn.lr_scheduler import CosineAnnealingWarmupLR # from colossalai.nn.lr_scheduler import CosineAnnealingWarmupLR
from colossalai.nn.optimizer import HybridAdam from colossalai.nn.optimizer import HybridAdam
from prefetch_generator import BackgroundGenerator from prefetch_generator import BackgroundGenerator
from pytorch_lightning import seed_everything from lightning.pytorch import seed_everything
from pytorch_lightning.trainer import Trainer from lightning.pytorch.trainer import Trainer
from pytorch_lightning.callbacks import ModelCheckpoint, Callback, LearningRateMonitor from lightning.pytorch.callbacks import ModelCheckpoint, Callback, LearningRateMonitor
from pytorch_lightning.utilities.rank_zero import rank_zero_only from lightning.pytorch.utilities.rank_zero import rank_zero_only
from pytorch_lightning.utilities import rank_zero_info from lightning.pytorch.utilities import rank_zero_info
from diffusers.models.unet_2d import UNet2DModel from diffusers.models.unet_2d import UNet2DModel
from clip.model import Bottleneck from clip.model import Bottleneck
@ -610,7 +610,7 @@ if __name__ == "__main__":
# default logger configs # default logger configs
default_logger_cfgs = { default_logger_cfgs = {
"wandb": { "wandb": {
"target": "pytorch_lightning.loggers.WandbLogger", "target": "lightning.pytorch.loggers.WandbLogger",
"params": { "params": {
"name": nowname, "name": nowname,
"save_dir": logdir, "save_dir": logdir,
@ -619,7 +619,7 @@ if __name__ == "__main__":
} }
}, },
"tensorboard":{ "tensorboard":{
"target": "pytorch_lightning.loggers.TensorBoardLogger", "target": "lightning.pytorch.loggers.TensorBoardLogger",
"params":{ "params":{
"save_dir": logdir, "save_dir": logdir,
"name": "diff_tb", "name": "diff_tb",
@ -642,7 +642,7 @@ if __name__ == "__main__":
print("Using strategy: {}".format(strategy_cfg["target"])) print("Using strategy: {}".format(strategy_cfg["target"]))
else: else:
strategy_cfg = { strategy_cfg = {
"target": "pytorch_lightning.strategies.DDPStrategy", "target": "lightning.pytorch.strategies.DDPStrategy",
"params": { "params": {
"find_unused_parameters": False "find_unused_parameters": False
} }
@ -654,7 +654,7 @@ if __name__ == "__main__":
# modelcheckpoint - use TrainResult/EvalResult(checkpoint_on=metric) to # modelcheckpoint - use TrainResult/EvalResult(checkpoint_on=metric) to
# specify which metric is used to determine best models # specify which metric is used to determine best models
default_modelckpt_cfg = { default_modelckpt_cfg = {
"target": "pytorch_lightning.callbacks.ModelCheckpoint", "target": "lightning.pytorch.callbacks.ModelCheckpoint",
"params": { "params": {
"dirpath": ckptdir, "dirpath": ckptdir,
"filename": "{epoch:06}", "filename": "{epoch:06}",
@ -722,7 +722,7 @@ if __name__ == "__main__":
'Caution: Saving checkpoints every n train steps without deleting. This might require some free space.') 'Caution: Saving checkpoints every n train steps without deleting. This might require some free space.')
default_metrics_over_trainsteps_ckpt_dict = { default_metrics_over_trainsteps_ckpt_dict = {
'metrics_over_trainsteps_checkpoint': 'metrics_over_trainsteps_checkpoint':
{"target": 'pytorch_lightning.callbacks.ModelCheckpoint', {"target": 'lightning.pytorch.callbacks.ModelCheckpoint',
'params': { 'params': {
"dirpath": os.path.join(ckptdir, 'trainstep_checkpoints'), "dirpath": os.path.join(ckptdir, 'trainstep_checkpoints'),
"filename": "{epoch:06}-{step:09}", "filename": "{epoch:06}-{step:09}",

1
examples/images/diffusion/requirements.txt
View File

@ -7,6 +7,7 @@ imageio==2.9.0
imageio-ffmpeg==0.4.2 imageio-ffmpeg==0.4.2
omegaconf==2.1.1 omegaconf==2.1.1
multiprocess multiprocess
lightning==1.8.1
test-tube>=0.7.5 test-tube>=0.7.5
streamlit>=0.73.1 streamlit>=0.73.1
einops==0.3.0 einops==0.3.0

2
examples/images/diffusion/scripts/img2img.py
View File

@ -13,7 +13,7 @@ from torchvision.utils import make_grid
from torch import autocast from torch import autocast
from contextlib import nullcontext from contextlib import nullcontext
import time import time
from pytorch_lightning import seed_everything from lightning.pytorch import seed_everything
from ldm.util import instantiate_from_config from ldm.util import instantiate_from_config
from ldm.models.diffusion.ddim import DDIMSampler from ldm.models.diffusion.ddim import DDIMSampler

2
examples/images/diffusion/scripts/txt2img.py
View File

@ -10,7 +10,7 @@ from itertools import islice
from einops import rearrange from einops import rearrange
from torchvision.utils import make_grid from torchvision.utils import make_grid
import time import time
from pytorch_lightning import seed_everything from lightning.pytorch import seed_everything
from torch import autocast from torch import autocast
from contextlib import contextmanager, nullcontext from contextlib import contextmanager, nullcontext