LMOps

 

LMOps is a research initiative on fundamental research and technology for building AI products w/ foundation models, especially on the general technology for enabling AI capabilities w/ LLMs and Generative AI models.

• Better Prompts: Automatic Prompt Optimization, Promptist, Extensible prompts, Universal prompt retrieval, LLM Retriever, In-Context Demonstration Selection
• Longer Context: Structured prompting, Length-Extrapolatable Transformers
• LLM Alignment: Alignment via LLM feedback
• LLM Accelerator (Faster Inference): Lossless Acceleration of LLMs
• LLM Customization: Adapt LLM to domains
• Fundamentals: Understanding In-Context Learning

Links

 

• microsoft/unilm: Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities
• microsoft/torchscale: Transformers at (any) Scale

News

 

• [Paper Release] Nov, 2023: In-Context Demonstration Selection with Cross Entropy Difference (EMNLP 2023)
• [Paper Release] Oct, 2023: Tuna: Instruction Tuning using Feedback from Large Language Models (EMNLP 2023)
• [Paper Release] Oct, 2023: Automatic Prompt Optimization with "Gradient Descent" and Beam Search (EMNLP 2023)
• [Paper Release] Oct, 2023: UPRISE: Universal Prompt Retrieval for Improving Zero-Shot Evaluation (EMNLP 2023)
• [Paper Release] July, 2023: Learning to Retrieve In-Context Examples for Large Language Models
• [Paper Release] April, 2023: Inference with Reference: Lossless Acceleration of Large Language Models
• [Paper Release] Dec, 2022: Why Can GPT Learn In-Context? Language Models Secretly Perform Finetuning as Meta Optimizers
• [Paper & Model & Demo Release] Dec, 2022: Optimizing Prompts for Text-to-Image Generation
• [Paper & Code Release] Dec, 2022: Structured Prompting: Scaling In-Context Learning to 1,000 Examples
• [Paper Release] Nov, 2022: Extensible Prompts for Language Models

Prompt Intelligence

 

Advanced technologies facilitating prompting language models.

Promptist: reinforcement learning for automatic prompt optimization

 

[Paper] Optimizing Prompts for Text-to-Image Generation

• Language models serve as a prompt interface that optimizes user input into model-preferred prompts.

• Learn a language model for automatic prompt optimization via reinforcement learning.

Structured Prompting: consume long-sequence prompts in an efficient way

 

[Paper] Structured Prompting: Scaling In-Context Learning to 1,000 Examples

• Example use cases:

1. Prepend (many) retrieved (long) documents as context in GPT.

2. Scale in-context learning to many demonstration examples.

X-Prompt: extensible prompts beyond NL for descriptive instructions

 

[Paper] Extensible Prompts for Language Models

• Extensible interface allowing prompting LLMs beyond natural language for fine-grain specifications

• Context-guided imaginary word learning for general usability

LLMA: LLM Accelerators

 

Accelerate LLM Inference with References

 

[Paper] Inference with Reference: Lossless Acceleration of Large Language Models

• Outputs of LLMs often have significant overlaps with some references (e.g., retrieved documents).

• LLMA losslessly accelerate the inference of LLMs by copying and verifying text spans from references into the LLM inputs.

• Applicable to important LLM scenarios such as retrieval-augmented generation and multi-turn conversations.

• Achieves 2~3 times speed-up without additional models.

Fundamental Understanding of LLMs

 

Understanding In-Context Learning

 

[Paper] Why Can GPT Learn In-Context? Language Models Secretly Perform Finetuning as Meta Optimizers

• According to the demonstration examples, GPT produces meta gradients for In-Context Learning (ICL) through forward computation. ICL works by applying these meta gradients to the model through attention.

• The meta optimization process of ICL shares a dual view with finetuning that explicitly updates the model parameters with back-propagated gradients.

• We can translate optimization algorithms (such as SGD with Momentum) to their corresponding Transformer architectures.

Hiring: aka.ms/GeneralAI

 

We are hiring at all levels (including FTE researchers and interns)! If you are interested in working with us on Foundation Models (aka large-scale pre-trained models) and AGI, NLP, MT, Speech, Document AI and Multimodal AI, please send your resume to fuwei@microsoft.com.