abacusai
Datasets
All datasets matching “abacusai”LongChat-Lines
Dataset Card for "LongChat-Lines"
This dataset is was used to evaluate the performance of model finetuned to operate on longer contexts. It is based on
a task template proposed by LMSys to evaluate attention to arbitrary points in the context. See the full details at
https;//github.com/abacusai/Long-Context.
WikiQA-Free_Form_QA
Dataset Card for "WikiQA-Free_Form_QA"
The WikiQA task is the task of answering a question based on the information given in a Wikipedia document. We have built upon the short answer format data in Google Natural Questions to construct our QA task. It is formatted as a document and a question. We ensure the answer to the question is a short answer which is either a single word or a small sentence directly cut pasted from the document. Having the task structured as such, we can… See the full description on the dataset page: https://huggingface.co/datasets/abacusai/WikiQA-Free_Form_QA.details_abacusai__MM-OV-bagel-DPO-34b-c1000-250
Dataset Card for Evaluation run of abacusai/MM-OV-bagel-DPO-34b-c1000-250
Dataset automatically created during the evaluation run of model abacusai/MM-OV-bagel-DPO-34b-c1000-250 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_abacusai__MM-OV-bagel-DPO-34b-c1000-250.MetaMathFewshot
A few-shot version of the MetaMath (https://huggingface.co/datasets/meta-math/MetaMathQA) dataset.
Each entry is formatted with 'question' and 'answer' keys. The 'question' key has a random number of query-answer pairs between 0 and 4 inclusive, before a final target query; the expected answer to this is stored in the content of 'answer'.
details_abacusai__MetaMath-bagel-34b-v0.2-c1500
Dataset Card for Evaluation run of abacusai/MetaMath-bagel-34b-v0.2-c1500
Dataset automatically created during the evaluation run of model abacusai/MetaMath-bagel-34b-v0.2-c1500 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_abacusai__MetaMath-bagel-34b-v0.2-c1500.MetaMath_DPO_FewShot
Dataset Card for "MetaMath_DPO_FewShot"
GSM8K \citep{cobbe2021training} is a dataset of diverse grade school maths word problems, which has been commonly adopted as a measure of the math and reasoning skills of LLMs.
The MetaMath dataset is an extension of the training set of GSM8K using data augmentation.
It is partitioned into queries and responses, where the query is a question involving mathematical calculation or reasoning, and the response is a logical series of steps and… See the full description on the dataset page: https://huggingface.co/datasets/abacusai/MetaMath_DPO_FewShot.