MACHINE LEARNING - AI Dynamics

Language Models Using Long Contexts: Analysis and Findings

By

–

07 July 2023 5h19

Lost in the Middle: How Language Models Use Long Contexts paper page: https://
huggingface.co/papers/2307.03
172
… While recent language models have the ability to take long contexts as input, relatively little is known about how well the language models use longer context. We analyze language

→ View original post on X — @_akhaliq

7 July 2023

Whisper-AT: Noise-Robust Speech Recognition and Audio Event Tagging

By

@_akhaliq

–

07 July 2023 5h00

Whisper-AT: Noise-Robust Automatic Speech Recognizers are Also Strong General Audio Event Taggers paper page: https://
huggingface.co/papers/2307.03
183
… In this paper, we focus on Whisper, a recent automatic speech recognition model trained with a massive 680k hour labeled speech corpus recorded

→ View original post on X — @_akhaliq

7 July 2023

VideoGLUE: Evaluating Foundation Models Video Understanding Capabilities

By

@_akhaliq

–

07 July 2023 4h20

GLUE: General Understanding Evaluation of Foundation Models paper page: https://
huggingface.co/papers/2307.03
166
… We evaluate existing foundation models video understanding capabilities using a carefully designed experiment protocol consisting of three hallmark tasks (action

→ View original post on X — @_akhaliq

7 July 2023

LongLLaMA: Large Language Model Supporting 256k Token Contexts

By

@_akhaliq

–

07 July 2023 3h52

TLDR LongLLaMA, a large language model capable of handling long contexts of 256k tokens or even more https://
huggingface.co/syzymon/long_l
lama_3b
…

→ View original post on X — @_akhaliq

7 July 2023

Focused Transformer: Contrastive Training for Context Scaling

By

@_akhaliq

–

07 July 2023 3h44

Focused Transformer: Contrastive Training for Context Scaling paper page: https://
huggingface.co/papers/2307.03
170
…
model: https://
huggingface.co/syzymon/long_l
lama_3b
… Large language models have an exceptional capability to incorporate new information in a contextual manner. However, the full potential of such an

→ View original post on X — @_akhaliq

7 July 2023

Training Models to Generate, Recognize, and Reframe Unhelpful Thoughts

By

@_akhaliq

–

07 July 2023 3h38

Training Models to Generate, Recognize, and Reframe Unhelpful Thoughts paper page: https://
huggingface.co/papers/2307.02
768
… Many cognitive approaches to well-being, such as recognizing and reframing unhelpful thoughts, have received considerable empirical support over the past decades, yet

→ View original post on X — @_akhaliq

7 July 2023

SkipDecode: Efficient LLM Inference with Autoregressive Skip Decoding

By

@_akhaliq

–

07 July 2023 3h30

SkipDecode: Autoregressive Skip Decoding with Batching and Caching for Efficient LLM Inference paper page: https://
huggingface.co/papers/2307.02
628
… Autoregressive large language models (LLMs) have made remarkable progress in various natural language generation tasks. However, they incur high

→ View original post on X — @_akhaliq

7 July 2023

mPLUG-DocOwl: Multimodal LLM for Document Understanding

By

@_akhaliq

–

07 July 2023 3h27

mPLUG-DocOwl: Modularized Multimodal Large Language Model for Document Understanding paper page: https://
huggingface.co/papers/2307.02
499
… Document understanding refers to automatically extract, analyze and comprehend information from various types of digital documents, such as a web page.

→ View original post on X — @_akhaliq

7 July 2023