jeevananandanne Goto Github PK

advfact

alignscore

AlignScore, a metric for factual consistency evaluation.

bartscore

BARTScore: Evaluating Generated Text as Text Generation

benetech-mga

2nd Place solution for Benetech - Making Graphs Accessible Kaggle Competition

bert_nli

A Natural Language Inference (NLI) model based on Transformers (BERT and ALBERT)

building_robust_recsys_dhs23

Talk: Building Robust and Scalable Recommendation Systems for Online Food Delivery

chain-of-verification

This repository implements the chain of verification paper by Meta AI

cheque-easy

Simplifying cheque processing for banks using Transformers

coco-bert-longformer

consistest

A conversational benchmark for factual consistency

correction_pipeline

cpmi

Mutual Information Predicts Hallucinations in Abstractive Summarization

deep_training

深度学习训练框架

dialsummeval

Resources for paper "DialSummEval: Revisiting summarization evaluation for dialogues"

endless-awsw

Angels with Scaly Wings Mod that allows for AI-enhanced narratives.

evalbase

base code for summary evaluation experiments

evaluating-llms

Creating the tools and data sets necessary to evaluate vulnerabilities in LLMs.

evaluation_matrix

evaluation matrix: bert_score bart_score factcc

factcc

Project for the AI Fellowship at Insight - Fact checking of text summarization models using Transformers.

factcc1

a easy way to directly get the factcc score of your generated summary

factccx

factedit

factsumm

FactSumm: Factual Consistency Scorer for Abstractive Summarization

factual-consistency-analysis-of-dialogs

A human annotated dataset that determines if neural generated responses are factually inconsistent with a knowledge snippet.

factual_clinicalsumm

fairseq-detect-hallucination

Detect hallucinated tokens for conditional sequence generation.

faithful-summarization-generation

faithful_summarization

faithfulness

A library to evaluate factual correctness of abstractive summaries.

feedback-prize

This is the solution for 2nd rank in Kaggle competition: Feedback Prize - Evaluating Student Writing.