-
Add training pipelines, consistency distillation scripts, and interactive dashboard server
-
Fix ONNX time conditioning and sinusoidal embedding parity with PyTorch
-
Update README.md
-
Update README.md
-
Update README.md
-
Link and badge live Hugging Face Space playground
-
Upload 540k subquery index embeddings and text for nearest neighbor decoding
-
Add training pipelines, consistency distillation scripts, and interactive dashboard server
-
Fix ONNX input/output feed tensor names
-
Release sub-millisecond 1-bit consistency retriever: ONNX, INT4 QAT checkpoint, and native C++ engine
-
Update .gitattributes with binary rules
-
Update best checkpoint with stabilized EDM log-normal schedule (val loss 0.6571)
-
Upload load_model.py with huggingface_hub
-
Upload data/diffusion_dataset_540k.pt with huggingface_hub
-
Upload data/fanouts_540k.jsonl with huggingface_hub
-
Upload diffusion_540k_best.pt with huggingface_hub
-
Upload README.md with huggingface_hub
-
initial commit
