GitLab

* Fix tokenization issue in BaseLM.loglikelihood

* Add a regression script

* Use entire non-continuation length as context

---------
Co-authored-by: Hailey Schoelkopf <65563625+haileyschoelkopf@users.noreply.github.com>

Add support for loading GPTQ models via AutoGPTQ

* allow for hf-causal to take dtype arg

* document this change

Add perplexity task on arbitrary JSON data

Add option to dump prompts and completions to a JSON file

Evaluation Against Portion of Benchmark Data

Add results of various models in json and md format

Create output path directory if necessary

* add xcopa dataset

* add xstory_cloze dataset and run pre-commit

* fix xcopa validation and test sets

* add xwinograd dataset

* add pawsx task

* add xnli task

* update task table with recently added tasks

* remove unused metrics from paws-x

* add mgsm task and fix gsm8k

* fix gsm8k until

* update task table

fix adaptive batch crash when there are no new requests

Set PAD token to EOS token

When `limit` is <1, limit represents the percentage of the total number of examples.
If it is >=1,  then it means the number of examples per task (only use this for testing).