matchmaker

matchmaker copied to clipboard

Hi! *tl;dr:* If it's a good idea, which `matchmaker` modules should I use to index and rank documents using ColBERT-like retriever with token-level vector index? We are trying to evaluate...

stefanik12

How do you suggest we can further train the per-trainined ColBert model form this repo on our domain specific dataset. Currently the repo only supports training the model from scratch...

ashutosh-dwivedi-e3502

Thank you for sharing codes. I wonder whether there is a typo in the implementation of KLDivTeacherList class. The implementation is ```python class KLDivTeacherList(nn.Module): def __init__(self): super(KLDivTeacherList, self).__init__() self.kl =...

jihyukkim-nlp

Hello, I am trying to get the pairwise distillation workflow run, and I'm following the instructions [here](https://github.com/sebastian-hofstaetter/matchmaker/blob/master/documentation/distillation_workflow_pairwise-supervision.md#distillation-workflow-pairwise-supervision) but it seems like `teacher-train-scorer.py` already expects there to be files in the...

paul-chelarescu

There is a line in file preprocessing/convert_formats/msmarco_doc_create_train_input_orig.py: `from matchmaker.evaluation.msmarco_eval import *` but there is no evaluation filefolder in matchmaker, could you be kind enough to release the code for this...

yhy-2000

Hi Sebastian, Will matchmaker support DistributedDataParallel? It is usually significantly faster than DataParallel. Best, Si

SunSiShining

allennlp.data.data_loaders.multiprocess_data_loader.WorkerError: worker raised ValueError("TextField's token_indexers have not been set.\nDid you forget to call DatasetReader.apply_token_indexers(instance) on your instance?\nIf apply_token_indexers() is being called but you're still seeing this error, it may not be implemented correctly.")

1

Hi, I'm running the TK model on msmarco-passage dataset. And when it starts to validate the model after 4000 steps, it show error like follows: > 2021-07-20 17:23:02,579 INFO [eval_model]...

haiahaiah

Pre-trained model

5

Hi, Are there pre-trained models available for download? Thanks in advance.

subhashiniy

The Colbert paper mentions normalising the query and document embeddings.

SonuDixit

issues on installing requirements

1

Hi, I'm experiencing some issues while trying to install all the dependencies. Do you have any updates on how to resolve this? Here is the error message: ``` AttributeError: module...

carlesoctav