Add GIMM-VFI support (NeurIPS 2024) with single-pass arbitrary-timestep interpolation

Integrates GIMM-VFI alongside existing BIM/EMA/SGM models. Key feature: generates
all intermediate frames in one forward pass (no recursive 2x passes needed for 4x/8x).

- Vendor gimm_vfi_arch/ from kijai/ComfyUI-GIMM-VFI with device fixes
- Two variants: RAFT-based (~80MB) and FlowFormer-based (~123MB)
- Auto-download checkpoints from HuggingFace (Kijai/GIMM-VFI_safetensors)
- Three new nodes: Load GIMM-VFI Model, GIMM-VFI Interpolate, GIMM-VFI Segment Interpolate
- single_pass toggle: True=arbitrary timestep (default), False=recursive like other models
- ds_factor parameter for high-res input downscaling

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

gimm_vfi_arch/__init__.py

@@ -0,0 +1,15 @@
+from .generalizable_INR.gimmvfi_r import GIMMVFI_R
+from .generalizable_INR.gimmvfi_f import GIMMVFI_F
+from .generalizable_INR.configs import GIMMVFIConfig
+from .generalizable_INR.raft.raft import RAFT as GIMM_RAFT
+from .generalizable_INR.flowformer.core.FlowFormer.LatentCostFormer.transformer import FlowFormer as GIMM_FlowFormer
+from .generalizable_INR.flowformer.configs.submission import get_cfg as gimm_get_flowformer_cfg
+from .utils.utils import InputPadder as GIMMInputPadder, RaftArgs as GIMMRaftArgs, easydict_to_dict
+from .generalizable_INR.modules.softsplat import objCudacache as gimm_softsplat_cache
+
+
+def clear_gimm_caches():
+    """Clear cached CUDA kernels and warp grids for GIMM-VFI."""
+    from .generalizable_INR.modules.fi_utils import backwarp_tenGrid
+    backwarp_tenGrid.clear()
+    gimm_softsplat_cache.clear()