search

chequanghuy / TwinLiteNet

MIT License
127 stars 32 forks source link

export:name 'net' is not defined #17

Open 1623021453 opened 3 months ago

1623021453 commented 3 months ago

python export.py error:model = net.Net() NameError: name 'net' is not defined

harrylal commented 3 months ago

Fix in the following pull request: #8 .

https://github.com/chequanghuy/TwinLiteNet/blob/e57dcfca3c6513e616f09aebe04b2b14fb79a427/export.py#L125 Replace the above line with model = Net()

1623021453 commented 3 months ago

after replace, new error: model.load_state_dict(torch.load(weights)) File "/home/user/anaconda3/envs/TwinLiteNet/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1223, in load_state_dict raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for TwinLiteNet: Missing key(s) in state_dict: "encoder.level1.conv.weight", "encoder.level1.bn.weight", "encoder.level1.bn.bias", "encoder.level1.bn.running_mean", "encoder.level1.bn.running_var", "encoder.level1.act.weight", "encoder.b1.conv.weight", "encoder.b1.bn.weight", "encoder.b1.bn.bias", "encoder.b1.bn.running_mean", "encoder.b1.bn.running_var", "encoder.b1.act.weight", "encoder.level2_0.c1.conv.weight", "encoder.level2_0.d1.conv.weight", "encoder.level2_0.d2.conv.weight", "encoder.level2_0.d4.conv.weight", "encoder.level2_0.d8.conv.weight", "encoder.level2_0.d16.conv.weight", "encoder.level2_0.bn.weight", "encoder.level2_0.bn.bias", "encoder.level2_0.bn.running_mean", "encoder.level2_0.bn.running_var", "encoder.level2_0.act.weight", "encoder.level2.0.c1.conv.weight",

harrylal commented 3 months ago

please view issue #11 for the fix

1623021453 commented 3 months ago

when I test with test_trt.py, the image has no results or wrong results caeb782d-4a20b7c4

huzkun commented 1 month ago

替换后，出现新错误： model.load_state_dict(torch.load(weights)) 文件“/home/user/anaconda3/envs/TwinLiteNet/lib/python3.8/site-packages/torch/nn/modules/module.py”，第 1223 行，在 load_state_dict 中 引发 RuntimeError('加载 state_dict 时出错，{}:\n\t{}'.format( RuntimeError: 加载 TwinLiteNet 的 state_dict 时出错： state_dict 中缺少键：“encoder.level1.conv.weight”、“encoder.level1.bn.weight”、“encoder.level1.bn.bias”、“encoder.level1.bn.running_mean”、“encoder.level1.bn.running_var”、“encoder.level1.act.weight”、“encoder.b1.conv.weight”、“encoder.b1.bn.weight”， “编码器.b1.bn.bias”、“编码器.b1.bn.running_mean”、“编码器.b1.bn.running_var”、“编码器.b1.act.weight”、“编码器.level2_0.c1.conv.weight”、“编码器.level2_0.d1.conv.weight”、“编码器.level2_0.d2.conv.weight”、“编码器.level2_0.d4.conv.weight”、“编码器.level2_0.d8.conv.weight”、“编码器.level2_0.d16.conv.weight”、“编码器.level2_0.bn.weight”、“编码器.level2_0.bn.bias”、“编码器.level2_0.bn.running_mean”、“编码器.level2_0.bn.running_var”、“编码器.level2_0.act.weight” “编码器.level2.0.c1.conv.权重”，

I also encountered this problem, did you solve it?