analysis/method_final.log

/workspace/ego6d_rag/method_final.py:14: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  fnet = Forecaster(cin=12, P=1).to(DEV); fnet.enc.load_state_dict(torch.load(AN / 'p3_fcx/shared_encoder.pt', map_location=DEV)); fnet.eval()
/workspace/ego6d_rag/method_final.py:15: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  cnet = CamFormer().to(DEV); cnet.load_state_dict(torch.load(AN / 'p3_cf/camformer.pt', map_location=DEV)); cnet.eval()
/opt/conda/envs/trajgaze/lib/python3.10/site-packages/torch/nn/modules/transformer.py:409: UserWarning: The PyTorch API of nested tensors is in prototype stage and will change in the near future. (Triggered internally at ../aten/src/ATen/NestedTensorImpl.cpp:178.)
  output = torch._nested_tensor_from_mask(output, src_key_padding_mask.logical_not(), mask_check=False)
venue   mass  soft oracle  proxy  gate->  use
Loc_18  3.88  3.64   2.16  0.540       soft
Loc_19  2.33  0.78   0.66  0.509       soft
Loc_20  1.67  1.50   0.59  0.511       soft
Loc_21  1.12  0.68   0.58  0.606       soft
Loc_22  4.50  3.79   0.84  0.276       mass
Loc_23  5.06  6.35   2.27  0.308       mass
Loc_24  3.10  3.44   0.81  0.281       mass
Loc_25  2.77  2.97   0.66  0.279       mass
Loc_26  4.04  2.38   0.65  0.409       soft
Loc_27  1.61  1.42   0.76  0.482       soft
Loc_28  5.31  7.71   1.16  0.359       mass
Loc_29  3.88  4.55   0.94  0.265       mass

LOCALIZATION macro:  mass-mean 3.27 | soft(ungated) 3.27 | PROXY-GATED(thr=0.409) 2.92 | oracle-gate 2.86 | per-window oracle 1.01
   refs: FCx top-1 3.31 | chance 3.95 | Method1/1b 3.4-3.7 (failed)

SEMANTICS macro (dual FCx|CamFormer probe):  room 0.456  actF1 0.369
   refs: FCx room .462/actF1 .282 | CamFormer room .453/actF1 .353