-
Notifications
You must be signed in to change notification settings - Fork 0
/
fine_tune_detr_Mongoose.py
181 lines (146 loc) · 6.94 KB
/
fine_tune_detr_Mongoose.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
#%%
root = "/jmain02/home/J2AD016/jjw02/jjs00-jjw02/dat"
import os
#os.environ['TRANSFORMERS_CACHE'] = root + "/TRANSFORMERS_CACHE"
os.environ['WANDB_MODE'] = 'offline'
os.environ['HF_DATASETS_OFFLINE'] = '1'
os.environ['TRANSFORMERS_OFFLINE'] = '1'
import torchvision
from transformers import DetrFeatureExtractor
import numpy as np
import os
from PIL import Image, ImageDraw
import pytorch_lightning as pl
from transformers import DetrConfig, DetrForObjectDetection, AutoConfig
import torch
import wandb
from pytorch_lightning import Trainer
from torch.utils.data import DataLoader
#%%
batch_size = 4
#name = "rudder_detr"
name = "mongoose_detr"
class CocoDetection(torchvision.datasets.CocoDetection):
def __init__(self, img_folder, feature_extractor, train=True):
ann_file = os.path.join(img_folder, "custom_train.json" if train else "custom_val.json")
super(CocoDetection, self).__init__(img_folder, ann_file)
self.feature_extractor = feature_extractor
def __getitem__(self, idx):
# read in PIL image and target in COCO format
img, target = super(CocoDetection, self).__getitem__(idx)
# preprocess image and target (converting target to DETR format, resizing + normalization of both image and target)
image_id = self.ids[idx]
target = {'image_id': image_id, 'annotations': target}
encoding = self.feature_extractor(images=img, annotations=target, return_tensors="pt")
pixel_values = encoding["pixel_values"].squeeze() # remove batch dimension
target = encoding["labels"][0] # remove batch dimension
return pixel_values, target
def collate_fn(batch):
pixel_values = [item[0] for item in batch]
feature_extractor = DetrFeatureExtractor.from_pretrained("facebook/detr-resnet-50")
encoding = feature_extractor.pad_and_create_pixel_mask(pixel_values, return_tensors="pt")
labels = [item[1] for item in batch]
batch = {}
batch['pixel_values'] = encoding['pixel_values']
batch['pixel_mask'] = encoding['pixel_mask']
batch['labels'] = labels
return batch
class Detr(pl.LightningModule):
def __init__(self, lr=1e-4, lr_backbone=1e-5, weight_decay=1e-4):
super().__init__()
# replace COCO classification head with custom head
config = AutoConfig.from_pretrained("/jmain02/home/J2AD016/jjw02/jjs00-jjw02/.cache/huggingface/hub/models--facebook--detr-resnet-50/snapshots/c783425425d573f30483efb0660bf6207deea991/config.json")
config.id2label = id2label
config.label2id = label2id
self.model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50", ignore_mismatched_sizes=True, config=config)
# see https://github.com/PyTorchLightning/pytorch-lightning/pull/1896
self.lr = lr
self.lr_backbone = lr_backbone
self.weight_decay = weight_decay
def forward(self, pixel_values, pixel_mask):
outputs = self.model(pixel_values=pixel_values, pixel_mask=pixel_mask)
return outputs
def common_step(self, batch, batch_idx):
pixel_values = batch["pixel_values"]
pixel_mask = batch["pixel_mask"]
labels = [{k: v.to(self.device) for k, v in t.items()} for t in batch["labels"]]
outputs = self.model(pixel_values=pixel_values, pixel_mask=pixel_mask, labels=labels)
loss = outputs.loss
loss_dict = outputs.loss_dict
return loss, loss_dict
def training_step(self, batch, batch_idx):
loss, loss_dict = self.common_step(batch, batch_idx)
# logs metrics for each training_step,
# and the average across the epoch
#self.log("training_loss", loss)
wandb.log({'train/loss' :loss})
for k,v in loss_dict.items():
#self.log("train_" + k, v.item())
wandb.log({'train/' + k : v.item()})
return {'loss': loss}
# def validation_step(self, batch, batch_idx):
# loss, loss_dict = self.common_step(batch, batch_idx)
# #self.log("validation_loss", loss)
# wandb.log({'val/loss' :loss})
# for k,v in loss_dict.items():
# #self.log("validation_" + k, v.item())
# wandb.log({'val/' + k : v.item()})
# return {'loss': loss}
def configure_optimizers(self):
param_dicts = [
{"params": [p for n, p in self.named_parameters() if "backbone" not in n and p.requires_grad]},
{
"params": [p for n, p in self.named_parameters() if "backbone" in n and p.requires_grad],
"lr": self.lr_backbone,
},
]
optimizer = torch.optim.AdamW(param_dicts, lr=self.lr,
weight_decay=self.weight_decay)
return optimizer
def train_dataloader(self):
return train_dataloader
# def val_dataloader(self):
# return val_dataloader
# %%
def main():
model = Detr(lr=1e-4, lr_backbone=1e-5, weight_decay=1e-4)
outputs = model(pixel_values=batch['pixel_values'], pixel_mask=batch['pixel_mask'])
outputs.logits.shape
trainer = Trainer(gpus=torch.cuda.device_count(), gradient_clip_val=0.1, default_root_dir=os.path.join(root, 'Mongoose', name), max_epochs=10000, log_every_n_steps=5)
trainer.fit(model)
#%%
wandb.init(project="Mongoose", entity="frankslab")
feature_extractor = DetrFeatureExtractor.from_pretrained("facebook/detr-resnet-50")
#train_dataset = CocoDetection(img_folder='/jmain02/home/J2AD016/jjw02/jjs00-jjw02/dat"/ElodeaProject/BB4_combined_split/train', feature_extractor=feature_extractor)
#val_dataset = CocoDetection(img_folder='/jmain02/home/J2AD016/jjw02/jjs00-jjw02/dat/ElodeaProject/BB4_combined_split/val', feature_extractor=feature_extractor, train=False)
train_dataset = CocoDetection(img_folder='/jmain02/home/J2AD016/jjw02/jjs00-jjw02/dat/Mongoose/data/train', feature_extractor=feature_extractor)
# %%
print("Number of training examples:", len(train_dataset))
#print("Number of validation examples:", len(val_dataset))
# %%
# image_ids = train_dataset.coco.getImgIds()
# image_id = image_ids[np.random.randint(0, len(image_ids))]
# print('Image n°{}'.format(image_id))
# image = train_dataset.coco.loadImgs(image_id)[0]
# image = Image.open(os.path.join('/local/scratch/jrs596/dat/ElodeaProject/BB3_combined_split/train/rudder', image['file_name']))
# annotations = train_dataset.coco.imgToAnns[image_id]
# draw = ImageDraw.Draw(image, "RGBA")
cats = train_dataset.coco.cats
id2label = {k: v['name'] for k,v in cats.items()}
label2id = {v:k for k,v in id2label.items()}
# for annotation in annotations:
# box = annotation['bbox']
# class_idx = annotation['category_id']
# x,y,w,h = tuple(box)
# draw.rectangle((x,y,x+w,y+h), outline='red', width=1)
# draw.text((x, y), id2label[class_idx], fill='white')
train_dataloader = DataLoader(train_dataset, collate_fn=collate_fn, batch_size=batch_size, shuffle=True, num_workers=os.cpu_count())
#val_dataloader = DataLoader(val_dataset, collate_fn=collate_fn, batch_size=batch_size, num_workers=os.cpu_count())
batch = next(iter(train_dataloader))
# %%
batch.keys()
pixel_values, target = train_dataset[0]
pixel_values.shape
# %%
if __name__ == "__main__":
main()