evaluates时出现错误，命令为python train.py --model ai85tinierssd --dataset cat --use-bias --confusion --evaluate --exp-load-weights-from ../ai8x-synthesis/trained/ai85-cat-qat8-q.pth.tar -8 --device MAX78000，
提示错误，不知道时训练时配置文件问题还是命令问题，解决不了！！！
Log file for this run: D:\MAX78000\AI\ai8x-training\logs\2023.11.20-223854\2023.11.20-223854.log
{'start_epoch': 10, 'weight_bits': 8}
=> loading checkpoint ../ai8x-synthesis/trained/ai85-cat-qat8-q.pth.tar
=> Checkpoint contents:
+----------------------+-------------+---------------+
| Key | Type | Value |
|----------------------+-------------+---------------|
| arch | str | ai85tinierssd |
| compression_sched | dict | |
| epoch | int | 199 |
| extras | dict | |
| optimizer_state_dict | dict | |
| optimizer_type | type | Adam |
| state_dict | OrderedDict | |
+----------------------+-------------+---------------+
=> Checkpoint['extras'] contents:
+-----------------+--------+---------------+
| Key | Type | Value |
|-----------------+--------+---------------|
| best_epoch | int | 199 |
| best_mAP | Tensor | |
| best_top1 | int | 0 |
| clipping_method | str | MAX_BIT_SHIFT |
| current_mAP | Tensor | |
| current_top1 | int | 0 |
+-----------------+--------+---------------+
Loaded compression schedule from checkpoint (epoch 199)
Traceback (most recent call last):
File "train.py", line 1835, in <module>
main()
File "train.py", line 376, in main
model = apputils.load_lean_checkpoint(model, args.load_model_path,
File "d:\max78000\ai\ai8x-training\distiller\distiller\apputils\checkpoint.py", line 91, in load_lean_checkpoint
return load_checkpoint(model, chkpt_file, model_device=model_device,
File "d:\max78000\ai\ai8x-training\distiller\distiller\apputils\checkpoint.py", line 233, in load_checkpoint
anomalous_keys = model.load_state_dict(checkpoint['state_dict'], strict)
File "D:\software\Anaconda3\envs\ai8x_training\lib\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 TinierSSD:
size mismatch for pred_convs.cl_fire8.op.weight: copying a param with shape torch.Size([12, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([8, 32, 3, 3]).
size mismatch for pred_convs.cl_fire8.op.bias: copying a param with shape torch.Size([12]) from checkpoint, the shape in current model is torch.Size([8]).
size mismatch for pred_convs.cl_fire9.op.weight: copying a param with shape torch.Size([12, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([8, 32, 3, 3]).
size mismatch for pred_convs.cl_fire9.op.bias: copying a param with shape torch.Size([12]) from checkpoint, the shape in current model is torch.Size([8]).
size mismatch for pred_convs.cl_fire10.op.weight: copying a param with shape torch.Size([12, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([8, 32, 3, 3]).
size mismatch for pred_convs.cl_fire10.op.bias: copying a param with shape torch.Size([12]) from checkpoint, the shape in current model is torch.Size([8]).
size mismatch for pred_convs.cl_conv12_2.op.weight: copying a param with shape torch.Size([12, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([8, 16, 3, 3]).
size mismatch for pred_convs.cl_conv12_2.op.bias: copying a param with shape torch.Size([12]) from checkpoint, the shape in current model is torch.Size([8]).
