Hugging Face's logo

Spaces:
henry000
/
YOLO
Running

App Files Community
YOLO / yolo

Commit History

🐛 [Fix] MixUp augmentation (#105) (#110)
a9e32de
unverified

Adam Kunák commited on

🔨 [FIX] Fixes memory leak (#83)
dea5a8a
unverified

Martin Cerman mcerman-enlite commited on

⬆️ [Update] dependency and the load cache function
699f2d2

henry000 commited on

⬆️ [Update] mAP from 0~1 to percentage
0efc848

henry000 commited on

🐛 [Fix] mul-GPU @ valid bugs, return mAP @ rk 0
a44a5c1

henry000 commited on

🐛 [Fix] #38 local mAP bugs, one pd match < 1 gt
aee6ac8

henry000 commited on

✨ [Add] mutli-GPU on validation mode
e0c8580

henry000 commited on

🐛 [Fix] LR schedule bugs, batchidx should increase
10420ed

henry000 commited on

🎨 [Fix] logging bugs, only log on rank0 gpu
ed7d166

henry000 commited on

👽️ [Update] cudaamp, most torch not support amp yet
360e5d7

henry000 commited on

🐛 [Fix] loss, avg loss if using ddp
3b24cc8

henry000 commited on

🚨 [Update] cuda.amp to amp, fix torch warning
9fc0d0e

henry000 commited on

🔀 [Merge] branch 'DATASET' into TRAIN
46b5efe

henry000 commited on

♻️ [Refactor] the dataloader import and default aug
8fc6449

henry000 commited on

🦺 [Add] a RemoveOutlier handle outside bbox
489f14b

henry000 commited on

✨ [Add] a RandomCrop transform, align paper
c627401

henry000 commited on

🚑️ [Fix] bbox bugs, choice 1 st bbox when draw
dc88787

henry000 commited on

🦺 [Fix] draw image bbox bugs, make it robust
47093ac

henry000 commited on

✨ [Add] Visualize image after validation, train
9c191b9

henry000 commited on

✨ [Add] TensorBoard support, finish #52
7f8fc3e

henry000 commited on

📝 [Update] wandb logger name
7e9db93

henry000 commited on

♻️ [Refactor] the tf board name, change lower case
f4a5047

henry000 commited on

♻️ [Refactor] create_model function
0ff0fd0

henry000 commited on

🔀 [Merge] branch 'DEPLOY' into TEST
5c8e313

henry000 commited on

🚚 [Rename] class, fix #49 issue, it's not repN
8a2b98c

henry000 commited on

🦺 [Update] load weight, can load trained weight
6972568

henry000 commited on

🐛 [Fix] #56 bugs, the using converter in ipynb
f1bfd74

henry000 commited on

🔀 [Merge] branch 'DATASET' into TEST
868c821

henry000 commited on

🔀 [Merge] branch 'MODEL' into TEST
9c96910

henry000 commited on

🚚 [Update] the class_num&class_list to dataset
3eb85fd

henry000 commited on

🔒️ 🔒️ [Update] dataset, can also load bbox GT
d2d621f

henry000 commited on

🔒️ [Update] safe load model, handle mismatch weight
4775b4c

henry000 commited on

🔨 [Remove] device in loss func, not sure correct
60c4943

henry000 commited on

💄 [Update] logging function, log only rank0 device
1132d27

henry000 commited on

🔀 [Merge] branch 'TRAIN' into TEST
ba6baa6

henry000 commited on

🔀 [Merge] branch 'MODEL' into TEST
877244a

henry000 commited on

🔀 [Merge] branch 'DEPLOY' into TEST
814608f

henry000 commited on

⚗️ [Add] a experiment, a bias init formula
592f75d

henry000 commited on

✏️ [Fix] typo in loading coco gt files
ce70a60

henry000 commited on

🎨 [Add] type hint for converter, Anc2Box/Vec2Box
c289952

henry000 commited on

🔨 [Update] Padding color, align with origin v9
27c7e17

henry000 commited on

🔨 [Update] AMP, unscale gradient and update model
192b787

henry000 commited on

⚡️ [Update] Optimizer, with more fine LR
5b9da41

henry000 commited on

🐛 [Fix] loss function calculation, detach predicts
94fae81

henry000 commited on

🔧 [Add] const random seed
598b827

henry000 commited on

🔨 [Add] wegith transform for v9seg model
f1585d3

henry000 commited on

✨ [Add] v9-c-segment model, inference WIP
00c86de

henry000 commited on

🔧 [Add] v9-c-seg config file
b679c1b

henry000 commited on

🐛 [Fix] bugs in onnx's output and video source
100c13d

henry000 commited on

🧐 [Add] TODO for training without Aux branch
e530629

henry000 commited on