Skip to content

Papers

[Paper Reading] Draft & Verify: Lossless Large Language Model Acceleration via Self-Speculative Decoding

Highlights of This Paper

  • Quantization, pruning, and distillation can also accelerate models, but come with issues like changes in output distribution compared to the original model, as well as the cost of retraining.
  • The original Speculative Decoding faces the issue of requiring additional memory to run the draft model, whereas Self-Speculative Decoding uses part of its own neural network as the draft model.
  • The Adaptive Draft-Exiting Mechanism can automatically adjust the number of tokens predicted by the draft model based on confidence score thresholds.
Read More »[Paper Reading] Draft & Verify: Lossless Large Language Model Acceleration via Self-Speculative Decoding

[Paper Reading] Lifting the Curse of Multilinguality by Pre-training Modular Transformers

Cross-lingual Modular (X-Mod) is an interesting language model architecture that modularizes the parameters for different languages as Module Units, allowing the model to use separate parameters when fine-tuning for a new language, thereby (comparatively) avoiding the problem of catastrophic forgetting.

Read More »[Paper Reading] Lifting the Curse of Multilinguality by Pre-training Modular Transformers

[Paper Reading] RAGAS: Automated Evaluation of Retrieval Augmented Generation

Introduction

The year 2023 witnessed an explosion of generative AI technologies, with a myriad of applications emerging across various domains. In the field of Natural Language Processing (NLP), Large Language Models (LLMs) stand out as one of the most significant advancements. By training LLMs effectively and reducing hallucinations, they can significantly reduce human effort across a wide range of tasks.

Read More »[Paper Reading] RAGAS: Automated Evaluation of Retrieval Augmented Generation

[Paper Reading] Mistral 7B

Introduction

Mistral 7B is a large language model (LLM) proposed on September 27, 2023, trained by the Mistral AI team, which also released its weights as open source. Interestingly, it uses the highly permissive Apache 2.0 license, unlike Llama 2, which has its own Llama license terms. Therefore, Mistral 7B is truly "open source" (Llama's license requires discussion with Meta AI when the service volume reaches 700 million).

Read More »[Paper Reading] Mistral 7B

[Paper Reading] Kangaroo: Lossless Self-Speculative Decoding via Double Early Exiting

Introduction

The accelerated framework is proposed by Huawei Noah's Ark Lab, it replaces the small model used in the original speculative decoding with the shallow sub-network of the large model. Additionally, it employs an extra-trained adapter and the model’s own decoding head to generate speculative tokens, which are then verified by the large model. The subsequent operations are quite similar to the original speculative decoding process.

Read More »[Paper Reading] Kangaroo: Lossless Self-Speculative Decoding via Double Early Exiting

[Paper Reading] Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection

Introduction

RAG-based LLM is a well-known architecture in current usage of Large Language Models (LLM). It involves "retrieval" to provide the model with prior knowledge that it lacks during training, enabling the model to answer questions in the context of specific information.

Read More »[Paper Reading] Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection