qing
d17677bade
- export_onnx: opset 17→18，dynamic_axes 改 dynamic_shapes 消除弃用警告 - predict: action_stats 改为 torch tensor，消除 NumPy 2.0 __array_wrap__ 警告 - train/pretrain: 每 epoch 保存 last.pt，新增 --resume 断点续训练 - model: MobileNetEncoder 下载失败时回退到已缓存权重 - _set.sh: 安装时预缓存 MobileNetV2 权重，失败给出明确指引 - test: 新增 TestResumeCheckpoint（2 用例），总计 58 测试全绿
37 lines
1.4 KiB
Bash
Executable File
37 lines
1.4 KiB
Bash
Executable File
#!/bin/bash
|
|
cd "$(dirname "$0")" || exit
|
|
python3 -m venv venv
|
|
source venv/bin/activate
|
|
|
|
# CPU-only PyTorch (no CUDA dependency)
|
|
# 如需 GPU 支持，注释掉下面一行并改用：pip install torch torchvision
|
|
pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu
|
|
|
|
# Remaining deps from Tsinghua mirror
|
|
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
|
|
|
|
# ONNX 导出依赖（torch>=2.12 的 torch.onnx.export 走 dynamo 路径，onnxscript 为硬依赖）
|
|
pip install onnxscript onnxruntime -i https://pypi.tuna.tsinghua.edu.cn/simple
|
|
|
|
pip check
|
|
mkdir -p checkpoints data
|
|
|
|
# Pre-download MobileNetV2 weights for offline fine-tuning (optional, skip if backbone=simple)
|
|
python3 -c "
|
|
import torch, sys, os
|
|
hub_dir = torch.hub.get_dir()
|
|
cache = os.path.join(hub_dir, 'checkpoints', 'mobilenet_v2-b0353104.pth')
|
|
if os.path.exists(cache):
|
|
print(f'MobileNetV2 weights already cached: {cache}')
|
|
sys.exit(0)
|
|
url = 'https://download.pytorch.org/models/mobilenet_v2-b0353104.pth'
|
|
try:
|
|
torch.hub.load_state_dict_from_url(url, model_dir=os.path.join(hub_dir, 'checkpoints'))
|
|
print('MobileNetV2 weights cached to: ' + cache)
|
|
except Exception as e:
|
|
print(f'WARNING: MobileNetV2 weights download failed: {e}')
|
|
print(' backbone=simple still works; for mobilenet, manually download from:')
|
|
print(f' {url}')
|
|
print(f' and place at: {cache}')
|
|
" || echo "WARNING: MobileNetV2 weight pre-download skipped"
|