Feat/row serialize (#62)

Eval/run evals (#56)

* feat: created a parallelized version of _get_downstream_task_results

* fix: fixed minor typos

* fix: adding comments

* fix: fixed to have less indent

* fix: added comments for parallelize

* fix: adding more comments and reorganization of code

* fix: sort results by query id after parallelization

* fix: convert to structured outputs

* feat: parallel processing for text2sql

* fix: added back a missing string

* fix: fixed missing dict key

* fix: testing parallelized generation, speed is stunning

* fix: fix minor bugs

* fix: removed unneeded print

* fix: removed unnecessary comments

* fix: minor fixes to run evals

* fix: add reminder for converting factver to parallelized version

* fix: remove unnecessary comments

* fix: removed commented out code

* fix: use global variable name for lm configs

add support for dummy table in dictionary format (#52)

Merge branch 'eval/run-evals' of github.com:target-benchmark/target i…

…nto eval/run-evals

Merge pull request #24 from target-benchmark/retrieval-result-persist…

…ence

Retrieval result persistence