mmyolo/configs/yolox/pose/yolox-pose_m_8xb32-300e-rtm...

15 lines
658 B
Python
Raw Permalink Normal View History

Support yolox-pose based on mmpose (#694) * add * reproduce map * add typehint and doc * format code * replace key * add ut * format * format * format code * fix ut * fix ut * fix comment * fix comment * fix comment * [WIP][Feature] Support yolov5-Ins training * fix comment * change data flow and fix loss_mask compute * align the data pipeline * remove albu gt mask key * support yolov5 ins inference * fix multi gpu test * align the post_process with v8 * support training * support training * code formatting * code formatting * Support pad_param type (#672) * add half_pad_param * fix default fast_test * fix loss weight compute * add models * add dataset1 * add dataset2 * add dataset3 * add configs * re commit __init__ * re commit __init__ * re commit * del local * add typo * del PoseToDetConverter and BBoxKeypoints * del local changes * fix mask rescale, add segment merge, fix segment2bbox * fix pipeline * add dataset * fix typo * add resize in mmyolo * fix typo * del local * del local changes * del local changes * fix dir name * fix dir name * add FilterAnnotations * fix typo * new config for yolox-pose * fix typo * fix typo * fix clip and fix mask init * del pose dataset changes * fix YOLOv5DetDataPreprocessor * del local file * fix typo * del init_cfg * simplify config * fix batch size * fix batch size * fix typo * code formatting * code formatting * code formatting * code formatting * fix bug for FilterAnnotations * simpler way for FilterAnnotations * update config * [Fix] fix load image from file * shorten eval time * fix typo * add large model * [Add] Add docs and more config * [Fix] config type and test_formatting * [Fix] fix yolov5-ins_m packdetinputs * hand rebase from yolov5-ins * use new PackDetInputs * rebase fix typo * add mapping table * fix typo * add weight * del typo * del typo * add results * install mmpose, Keypoints note, context manager, predict, ota rename * fix test * add unittest for pose_sim_ota_assigner and yolox_head * add unittest for pose_sim_ota_assigner and yolox_head * fix typo --------- Co-authored-by: Nioolek <379319054@qq.com> Co-authored-by: josonchan <josonchan1998@163.com> Co-authored-by: Nioolek <40284075+Nioolek@users.noreply.github.com> Co-authored-by: huanghaian <huanghaian@sensetime.com>
2023-05-15 10:58:25 +08:00
_base_ = ['./yolox-pose_s_8xb32-300e-rtmdet-hyp_coco.py']
load_from = 'https://download.openmmlab.com/mmyolo/v0/yolox/yolox_m_fast_8xb32-300e-rtmdet-hyp_coco/yolox_m_fast_8xb32-300e-rtmdet-hyp_coco_20230210_144328-e657e182.pth' # noqa
# ========================modified parameters======================
deepen_factor = 0.67
widen_factor = 0.75
# =======================Unmodified in most cases==================
# model settings
model = dict(
backbone=dict(deepen_factor=deepen_factor, widen_factor=widen_factor),
neck=dict(deepen_factor=deepen_factor, widen_factor=widen_factor),
bbox_head=dict(head_module=dict(widen_factor=widen_factor)))