metadata
library_name: pytorch
tags:
- table-retrieval
- embedding-adapter
- centroid-adapter
license: mit
Centroid Adapter — bge
Lightweight BottleneckResidualAdapter trained on top of bge embeddings to produce representation-invariant table embeddings.
Architecture
z = e + α · Up( Dropout( GELU( Down( LN(e) ) ) ) )
| Hyperparameter | Value |
|---|---|
Embedding dim d |
1024 |
Bottleneck rank r |
512 |
Residual scale α |
0.01 |
| Use bias | True |
Trained on: WTQ, WIKISQL
Usage
import torch
from huggingface_hub import hf_hub_download
import json
# --- option A: use the from_pretrained helper in this repo ---
# (copy BottleneckResidualAdapter + from_pretrained from push_to_hub.py)
adapter = BottleneckResidualAdapter.from_pretrained("KBhandari11/centroid-adapter-subset-bge")
e = torch.randn(1, 1024) # your backbone embedding
z = adapter(e) # representation-invariant embedding
# --- option B: hf_hub_download one-liner ---
from safetensors.torch import load_file
weights_path = hf_hub_download("KBhandari11/centroid-adapter-subset-bge", "model.safetensors")
cfg_path = hf_hub_download("KBhandari11/centroid-adapter-subset-bge", "config.json")
with open(cfg_path) as f:
cfg = json.load(f)
adapter = BottleneckResidualAdapter(**cfg)
adapter.load_state_dict(load_file(weights_path))
adapter.eval()
Research Paper
Improving Robustness of Tabular Retrieval via Representational Stability