Files
HSAP/algorithms/lane_ufld/code.embedded.bak/UFLD/TRAIN_ENV_CPU.md
Chengfang Lu e72bc061c5 feat: HSAP platform v2 — modular navigation, quality review, audit log, world model simulation
Major changes:
- New frontend (platform/web/): Vite + React 18 + TypeScript + Tailwind
- 4-module navigation: 数据送标 / 模型管理 / 车队管理 / 系统管理
- Data catalog with charts (DMS/ADAS/Lane 3-tab view)
- Quality review workflow (标注质检): Good/Fine/Bad scoring with auto-advance
- Audit enhancements: batch operations, rejection categories, Feishu notifications
- Operation audit log (操作日志)
- World model simulation studio (仿真工坊)
- Dataset version management with snapshots and diff
- ADAS 7-class dataset integration (138K images organized + compressed)
- User management with Feishu integration and pagination
- CRUD/search/filter on all pages, card layout redesign
- PIL-optimized image overlay rendering
- Auto-snapshot on build, in_review workflow stage
- Removed embedded algorithm code (now in workspace)
2026-06-03 11:40:21 +08:00

1.6 KiB
Raw Blame History

lane_light + CPU PyTorchUFLD 训练)

1. 激活环境

source /home/chengfanglu/miniconda3/etc/profile.d/conda.sh
conda activate lane_light

2. 已安装依赖(lane_light

  • torch / torchvisionCPU 轮子,来自 https://download.pytorch.org/whl/cpu
  • opencv-python, tqdm, tensorboard, addict, scikit-learn, pathspec(与 requirements.txt 对齐;sklearn 包名在 pip 中为 scikit-learn

自检:

python -c "import torch; print('torch', torch.__version__, 'cuda=', torch.cuda.is_available())"

3. 数据与配置

  • 默认数据根仍指向 lane0_reorganized/lane_training_pack(见 configs/mufld_lane_culane.py)。
  • CPU 建议使用 configs/mufld_lane_culane_cpu.pybatch_size=4,学习率与 warmup 已按 batch 相对 16 做了粗略缩放)。内存不够可改配置或命令行覆盖:
cd /home/chengfanglu/DATA/BK2/UFLD
python train.py configs/mufld_lane_culane_cpu.py --batch_size 2

4. 运行训练

cd /home/chengfanglu/DATA/BK2/UFLD
python train.py configs/mufld_lane_culane_cpu.py

说明:

  • train.py 已改为在 无 CUDA 时使用 cpu;原仓库中写死的 CUDA_VISIBLE_DEVICES=1,2.cuda() 已去掉,避免 CPU 机直接报错。
  • 首次 pretrained=True 会下载 ResNet 骨干权重,需联网。
  • CPU 训练很慢,建议先用小 epoch / 小 batch_size 做通路测试。

5. 可选DataLoader num_workers

当前 data/dataloader.pynum_workers=8。若 CPU 内存紧张或不想多进程读盘,可自行把该值改小(例如 02)。