the train method has shown the parameter like this
LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]
| Name | Type | Params
0 | model | NanoDetPlus | 7.7 M
1 | avg_model | NanoDetPlus | 7.7 M
15.5 M Trainable params
0 Non-trainable params
15.5 M Total params
61.915 Total estimated model params size (MB)
but what I saw in readme
the parameter is 2.44M, is there any wrong between them?
My config is copied by nanodet-plus-m-1.5x-416.yaml
the train method has shown the parameter like this LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]
| Name | Type | Params
0 | model | NanoDetPlus | 7.7 M 1 | avg_model | NanoDetPlus | 7.7 M
15.5 M Trainable params 0 Non-trainable params 15.5 M Total params 61.915 Total estimated model params size (MB) but what I saw in readme the parameter is 2.44M, is there any wrong between them? My config is copied by nanodet-plus-m-1.5x-416.yaml