Can you provide the Train/Validation/Test splits and the required settings for reproducing your work.
can you also share your best model that provides the results in the paper? (did you use the best or final checkpoint?)
Missing key(s) in state_dict: "vit.stages.0.0.mamba.A_b_log", "vit.stages.0.0.mamba.D_b", "vit.stages.0.0.mamba.A_s_log", "vit.stages.0.0.mamba.D_s", "vit.stages.0.0.mamba.conv1d_b.weight", "vit.stages.0.0.mamba.conv1d_b.bias", "vit.stages.0.0.mamba.x_proj_b.weight", "vit.stages.0.0.mamba.dt_proj_b.weight", "vit.stages.0.0.mamba.dt_proj_b.bias", "vit.stages.0.0.mamba.conv1d_s.weight", "vit.stages.0.0.mamba.conv1d_s.bias", "vit.stages.0.0.mamba.x_proj_s.weight", "vit.stages.0.0.mamba.dt_proj_s.weight", "vit.stages.0.0.mamba.dt_proj_s.bias", "vit.stages.0.1.mamba.A_b_log", "vit.stages.0.1.mamba.D_b", "vit.stages.0.1.mamba.A_s_log", "vit.stages.0.1.mamba.D_s", "vit.stages.0.1.mamba.conv1d_b.weight", "vit.stages.0.1.mamba.conv1d_b.bias", "vit.stages.0.1.mamba.x_proj_b.weight", "vit.stages.0.1.mamba.dt_proj_b.weight", "vit.stages.0.1.mamba.dt_proj_b.bias", "vit.stages.0.1.mamba.conv1d_s.weight", "vit.stages.0.1.mamba.conv1d_s.bias", "vit.stages.0.1.mamba.x_proj_s.weight", "vit.stages.0.1.mamba.dt_proj_s.weight", "vit.stages.0.1.mamba.dt_proj_s.bias", "vit.stages.1.0.mamba.A_b_log", "vit.stages.1.0.mamba.D_b", "vit.stages.1.0.mamba.A_s_log", "vit.stages.1.0.mamba.D_s", "vit.stages.1.0.mamba.conv1d_b.weight", "vit.stages.1.0.mamba.conv1d_b.bias", "vit.stages.1.0.mamba.x_proj_b.weight", "vit.stages.1.0.mamba.dt_proj_b.weight", "vit.stages.1.0.mamba.dt_proj_b.bias", "vit.stages.1.0.mamba.conv1d_s.weight", "vit.stages.1.0.mamba.conv1d_s.bias", "vit.stages.1.0.mamba.x_proj_s.weight", "vit.stages.1.0.mamba.dt_proj_s.weight", "vit.stages.1.0.mamba.dt_proj_s.bias", "vit.stages.1.1.mamba.A_b_log", "vit.stages.1.1.mamba.D_b", "vit.stages.1.1.mamba.A_s_log", "vit.stages.1.1.mamba.D_s", "vit.stages.1.1.mamba.conv1d_b.weight", "vit.stages.1.1.mamba.conv1d_b.bias", "vit.stages.1.1.mamba.x_proj_b.weight", "vit.stages.1.1.mamba.dt_proj_b.weight", "vit.stages.1.1.mamba.dt_proj_b.bias", "vit.stages.1.1.mamba.conv1d_s.weight", "vit.stages.1.1.mamba.conv1d_s.bias", "vit.stages.1.1.mamba.x_proj_s.weight", "vit.stages.1.1.mamba.dt_proj_s.weight", "vit.stages.1.1.mamba.dt_proj_s.bias", "vit.stages.2.0.mamba.A_b_log", "vit.stages.2.0.mamba.D_b", "vit.stages.2.0.mamba.A_s_log", "vit.stages.2.0.mamba.D_s", "vit.stages.2.0.mamba.conv1d_b.weight", "vit.stages.2.0.mamba.conv1d_b.bias", "vit.stages.2.0.mamba.x_proj_b.weight", "vit.stages.2.0.mamba.dt_proj_b.weight", "vit.stages.2.0.mamba.dt_proj_b.bias", "vit.stages.2.0.mamba.conv1d_s.weight", "vit.stages.2.0.mamba.conv1d_s.bias", "vit.stages.2.0.mamba.x_proj_s.weight", "vit.stages.2.0.mamba.dt_proj_s.weight", "vit.stages.2.0.mamba.dt_proj_s.bias", "vit.stages.2.1.mamba.A_b_log", "vit.stages.2.1.mamba.D_b", "vit.stages.2.1.mamba.A_s_log", "vit.stages.2.1.mamba.D_s", "vit.stages.2.1.mamba.conv1d_b.weight", "vit.stages.2.1.mamba.conv1d_b.bias", "vit.stages.2.1.mamba.x_proj_b.weight", "vit.stages.2.1.mamba.dt_proj_b.weight", "vit.stages.2.1.mamba.dt_proj_b.bias", "vit.stages.2.1.mamba.conv1d_s.weight", "vit.stages.2.1.mamba.conv1d_s.bias", "vit.stages.2.1.mamba.x_proj_s.weight", "vit.stages.2.1.mamba.dt_proj_s.weight", "vit.stages.2.1.mamba.dt_proj_s.bias", "vit.stages.3.0.mamba.A_b_log", "vit.stages.3.0.mamba.D_b", "vit.stages.3.0.mamba.A_s_log", "vit.stages.3.0.mamba.D_s", "vit.stages.3.0.mamba.conv1d_b.weight", "vit.stages.3.0.mamba.conv1d_b.bias", "vit.stages.3.0.mamba.x_proj_b.weight", "vit.stages.3.0.mamba.dt_proj_b.weight", "vit.stages.3.0.mamba.dt_proj_b.bias", "vit.stages.3.0.mamba.conv1d_s.weight", "vit.stages.3.0.mamba.conv1d_s.bias", "vit.stages.3.0.mamba.x_proj_s.weight", "vit.stages.3.0.mamba.dt_proj_s.weight", "vit.stages.3.0.mamba.dt_proj_s.bias", "vit.stages.3.1.mamba.A_b_log", "vit.stages.3.1.mamba.D_b", "vit.stages.3.1.mamba.A_s_log", "vit.stages.3.1.mamba.D_s", "vit.stages.3.1.mamba.conv1d_b.weight", "vit.stages.3.1.mamba.conv1d_b.bias", "vit.stages.3.1.mamba.x_proj_b.weight", "vit.stages.3.1.mamba.dt_proj_b.weight", "vit.stages.3.1.mamba.dt_proj_b.bias", "vit.stages.3.1.mamba.conv1d_s.weight", "vit.stages.3.1.mamba.conv1d_s.bias", "vit.stages.3.1.mamba.x_proj_s.weight", "vit.stages.3.1.mamba.dt_proj_s.weight", "vit.stages.3.1.mamba.dt_proj_s.bias".
Hello,
i hope that you can assist me,
Can you provide the Train/Validation/Test splits and the required settings for reproducing your work.
can you also share your best model that provides the results in the paper? (did you use the best or final checkpoint?)
i've tried to download the final check point you have shared here: #55
but i couldn't load it. does it use the TOM? because i get the following error:
Thanks