Instructions to use webis/colbert with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Lightning IR
How to use webis/colbert with Lightning IR:
#install from https://github.com/webis-de/lightning-ir from lightning_ir import BiEncoderModule model = BiEncoderModule("webis/colbert") model.score("query", ["doc1", "doc2", "doc3"]) - Notebooks
- Google Colab
- Kaggle
File size: 624 Bytes
bb86129 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | trainer:
logger: false
callbacks:
- class_path: lightning_ir.IndexCallback
init_args:
index_dir: ./index
index_config:
class_path: FaissIVFPQIndexConfig
init_args:
num_centroids: 262144
num_subquantizers: 16
n_bits: 8
model:
class_path: lightning_ir.BiEncoderModule
init_args:
model_name_or_path: webis/bert-bi-encoder
data:
class_path: lightning_ir.LightningIRDataModule
init_args:
num_workers: 1
inference_batch_size: 256
inference_datasets:
- class_path: DocDataset
init_args:
doc_dataset: msmarco-passage |