Add initial implementation of repro matrix for MS MARCO v1 passage

[lintool]

The goal is a repo that will automatically generate the results table in the SIGIR 2022 resource paper. This is my initial implementation.

After running python scripts/run_all_msmarco_v1_passage.py, the following is automatically generated:

                                                 TREC 2019                TREC 2020            MS MARCO dev
                                             MAP nDCG@10    R@1K       MAP nDCG@10    R@1K    MRR@10    R@1K
                                          ----------------------    ----------------------    --------------
BM25 (k1=0.82, b=0.68)                    0.2903  0.4973  0.7450    0.2876  0.4876  0.8031    0.1875  0.8573
BM25+RM3 (k1=0.82, b=0.68)                0.3377  0.5231  0.7792    0.3056  0.4808  0.8286    0.1668  0.8687
BM25 w/ doc2query-T5 (k1=2.18, b=0.86)    0.4046  0.6336  0.8134    0.4171  0.6265  0.8393    0.2816  0.9506

BM25 (k1=0.9, b=0.4)                      0.3013  0.5058  0.7501    0.2856  0.4796  0.7863    0.1840  0.8526
BM25+RM3 (k1=0.9, b=0.4)                  0.3390  0.5180  0.7998    0.3019  0.4821  0.8217    0.1564  0.8606
BM25 w/ doc2query-T5 (k1=0.9, b=0.4)      0.4034  0.6417  0.8310    0.4074  0.6187  0.8452    0.2723  0.9470

uniCOIL (w/ doc2query-T5) - pre-encoded   0.4612  0.7024  0.8292    0.4430  0.6745  0.8430    0.3516  0.9582
uniCOIL (w/ doc2query-T5) - otf           0.4617  0.7027  0.8291    0.4429  0.6745  0.8433    0.3509  0.9581

For reference, here is the results table in the paper:

[ronakice]

LGTM!

[lintool]

We can even incorporate dense retrieval:

                                                 TREC 2019                TREC 2020            MS MARCO dev
                                             MAP nDCG@10    R@1K       MAP nDCG@10    R@1K    MRR@10    R@1K
                                          ----------------------    ----------------------    --------------
BM25 (k1=0.82, b=0.68)                    0.2903  0.4973  0.7450    0.2876  0.4876  0.8031    0.1875  0.8573
BM25+RM3 (k1=0.82, b=0.68)                0.3377  0.5231  0.7792    0.3056  0.4808  0.8286    0.1668  0.8687
BM25 w/ doc2query-T5 (k1=2.18, b=0.86)    0.4046  0.6336  0.8134    0.4171  0.6265  0.8393    0.2816  0.9506

BM25 (k1=0.9, b=0.4)                      0.3013  0.5058  0.7501    0.2856  0.4796  0.7863    0.1840  0.8526
BM25+RM3 (k1=0.9, b=0.4)                  0.3390  0.5180  0.7998    0.3019  0.4821  0.8217    0.1564  0.8606
BM25 w/ doc2query-T5 (k1=0.9, b=0.4)      0.4034  0.6417  0.8310    0.4074  0.6187  0.8452    0.2723  0.9470

uniCOIL (w/ doc2query-T5) - pre-encoded   0.4612  0.7024  0.8292    0.4430  0.6745  0.8430    0.3516  0.9582
uniCOIL (w/ doc2query-T5) - otf           0.4617  0.7027  0.8291    0.4429  0.6745  0.8433    0.3509  0.9581

TCT_ColBERT-V2-HN+ - pre-encoded          0.0000  0.0000  0.0000    0.0000  0.0000  0.0000    0.3584  0.9695
TCT_ColBERT-V2-HN+ - otf                  0.4469  0.7204  0.8261    0.4754  0.6882  0.8429    0.3584  0.9695