Views
No views yet
quantized_video/, so the repo stays easy to move around and use as a single package.
The bundled loader keeps the original V-JEPA2 backbone id visible to TRIBEv2 consumers and swaps in the embedded quantized weights from quantized_video/ at load time.
All presets in this family keep a ViT-G-compatible video feature interface so the untouched TRIBEv2 checkpoint remains dimensionally compatible.facebook/vjepa2-vitg-fpc64-2562.0 Hz8defaultbest.ckptquantized_video/load_quantized_tribev2.py that injects the embedded quantized video branchpip install "torch>=2.11" transformers huggingface_hub pyyaml torchao tribev2TribeModel.from_pretrained(...) directly on the repo.1from pathlib import Path
2from load_quantized_tribev2 import load_quantized_tribev2
3
4repo_dir = Path(".")
5model = load_quantized_tribev2(repo_dir, device="auto")
6print(type(model).__name__)load_quantized_tribev2() keeps data.video_feature.image.model_name aligned with facebook/vjepa2-vitg-fpc64-256 for compatibility and injects the embedded quantized weights from quantized_video/.facebook/tribev2facebook/vjepa2-vitg-fpc64-256