mirrors
/
fast-reid
mirror of https://github.com/JDAI-CV/fast-reid.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
fast-reid/configs/Base-bagtricks.yml

77 lines
1.1 KiB
YAML
Raw Normal View History

style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
MODEL:
update fastreid V1.0
2021-01-18 11:36:38 +08:00
META_ARCHITECTURE: Baseline
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
BACKBONE:
update fastreid V1.0
2021-01-18 11:36:38 +08:00
NAME: build_resnet_backbone
NORM: BN
DEPTH: 50x
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
LAST_STRIDE: 1
change config file
2020-09-10 10:58:56 +08:00
FEAT_DIM: 2048
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
WITH_IBN: False
PRETRAIN: True
HEADS:
update fastreid V1.0
2021-01-18 11:36:38 +08:00
NAME: EmbeddingHead
NORM: BN
change config file
2020-09-10 10:58:56 +08:00
WITH_BNNECK: True
bugfix & merge classification transforms (#448) Summary: change heads definition in project and config file, merge classification transforms into default transforms
2021-03-30 15:47:14 +08:00
POOL_LAYER: GlobalAvgPool
update fastreid V1.0
2021-01-18 11:36:38 +08:00
NECK_FEAT: before
add configurable decorator & linear loss decouple (#441) Summary: Add configurable decorator which can call `Baseline` with `Baseline(cfg)` or `Baseline(cfg, heads=heads, ...)` Decouple linear and loss computation for partial-fc support. Reviewed By: l1aoxingyu
2021-03-23 12:10:06 +08:00
CLS_LAYER: Linear
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
LOSSES:
update all config files
2020-07-06 16:48:14 +08:00
NAME: ("CrossEntropyLoss", "TripletLoss",)
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
CE:
EPSILON: 0.1
chore(configs): update all training config
2020-05-01 09:04:51 +08:00
SCALE: 1.
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
TRI:
chore(configs): update Market1501 training config
2020-04-29 21:06:57 +08:00
MARGIN: 0.3
HARD_MINING: True
update all config files
2020-07-06 16:48:14 +08:00
NORM_FEAT: False
chore(configs): update all training config
2020-05-01 09:04:51 +08:00
SCALE: 1.
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
INPUT:
update fastreid V1.0
2021-01-18 11:36:38 +08:00
SIZE_TRAIN: [ 256, 128 ]
SIZE_TEST: [ 256, 128 ]
bugfix for attribute project (#450) Summary: refactor sample weight in attribute recognition; change all options to False in defaults.py and modify yaml files
2021-03-31 17:07:19 +08:00
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
REA:
ENABLED: True
PROB: 0.5
bugfix for attribute project (#450) Summary: refactor sample weight in attribute recognition; change all options to False in defaults.py and modify yaml files
2021-03-31 17:07:19 +08:00
FLIP:
ENABLED: True
support classification in fastreid (#443) Summary: support classification and refactor build_dataloader which can support explicit parameters passing
2021-03-26 20:17:39 +08:00
PADDING:
ENABLED: True
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
DATALOADER:
support classification in fastreid (#443) Summary: support classification and refactor build_dataloader which can support explicit parameters passing
2021-03-26 20:17:39 +08:00
SAMPLER_TRAIN: NaiveIdentitySampler
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
NUM_INSTANCE: 4
update all config files
2020-07-06 16:48:14 +08:00
NUM_WORKERS: 8
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
SOLVER:
support classification in fastreid (#443) Summary: support classification and refactor build_dataloader which can support explicit parameters passing
2021-03-26 20:17:39 +08:00
AMP:
ENABLED: True
update fastreid V1.0
2021-01-18 11:36:38 +08:00
OPT: Adam
v0.3 update Summary: 1. change DPP training in apex way; 2. make warmup scheduler by iter and lr scheduler by epoch; 3. replace random erasing with torchvision implementation; 4. naming modification in config file
2020-12-07 14:19:20 +08:00
MAX_EPOCH: 120
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
BASE_LR: 0.00035
WEIGHT_DECAY: 0.0005
Align bagtricks config with v1.2
2021-06-01 11:51:22 +08:00
WEIGHT_DECAY_NORM: 0.0005
chore(configs): update Market1501 training config
2020-04-29 21:06:57 +08:00
IMS_PER_BATCH: 64
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
update fastreid V1.0
2021-01-18 11:36:38 +08:00
SCHED: MultiStepLR
STEPS: [ 40, 90 ]
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
GAMMA: 0.1
v0.3 update Summary: 1. change DPP training in apex way; 2. make warmup scheduler by iter and lr scheduler by epoch; 3. replace random erasing with torchvision implementation; 4. naming modification in config file
2020-12-07 14:19:20 +08:00
WARMUP_FACTOR: 0.1
make lr warmup by iter Summary: change warmup way by iter not by epoch, which will make it more flexible when training small epochs
2021-01-22 11:17:21 +08:00
WARMUP_ITERS: 2000
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
v0.3 update Summary: 1. change DPP training in apex way; 2. make warmup scheduler by iter and lr scheduler by epoch; 3. replace random erasing with torchvision implementation; 4. naming modification in config file
2020-12-07 14:19:20 +08:00
CHECKPOINT_PERIOD: 30
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
TEST:
update all config files
2020-07-06 16:48:14 +08:00
EVAL_PERIOD: 30
IMS_PER_BATCH: 128
style(configs): put all config files together put all config files into one place for easily control, and add tools for put train_net.py which almost the same in different projects
2020-04-29 16:18:54 +08:00
CUDNN_BENCHMARK: True