Tong Guo

Predictions For Pre-training Language Models

2020 · arXiv (Cornell University)

Language model pre-training has proven to be useful in many language understanding tasks. In this paper, we investigate whether it is still helpful to add the self-training method in the pre-training step and the fine-tuning …

A Comprehensive Comparison of Pre-training Language Models

2021

<p>Recently, the development of pre-trained language models has brought natural language processing (NLP) tasks to the new state-of-the-art. In this paper we explore the efficiency of various pre-trained language models. We pre-train a list of …

Re-Label Is All You Need

2022

In industry deep learning application, our manually labeled data has a certain number of noisy data. To solve this problem and achieve more than 90 score in dev dataset, we present a simple method to …

A Comprehensive Exploration of Item Recommendation Text Ranking

2022

In this paper, we study the personalized ranking problem for item’s recommendation text. Item’s recommendation text is the sentence that describe the item highlights for user decision (such as buy or click). The recommendation texts …

Re-Label By Data Pattern For Knowledge Driven Deep Learning

2022

<p>In industry deep learning application, we should fix the badcase by human evaluation after we achieve more than 95% accuracy at dev dataset. The badcase reason is from the wrong rule/knowledge of human labeling and …

Re-Label By Data Pattern Is All You Need For Knowledge Driven Deep Learning

2022

In industry deep learning application, we should fix the badcase by human evaluation after we achieve more than 95% accuracy at dev dataset. The badcase reason is from the wrong rule/knowledge of human labeling and …

Answering High-precision Problems for LLMs by Combining Text2code

2024

The current large language models (LLMs) mainly lacks this capability: answering high-precision questions/prompts. LLMs is actually a powerful fuzzy memory system that makes it difficult to answer high-precision questions. The results of code execution is …