Large Language Models (LLMs)

Articles

1. GPT4 can improve itself

2. Lil Weng - Prompt Engineering

3. Chip Huyen - Building LLM applicaations for production

4. How to generate text using different decoding methods for language generation with transformers

5. (great) a gentle intro to LLMs and Langchain

6. LLMs can explain NN of other LLMs by OpenAI

7. Fine tuning LLMs

8. Model size vs Computer overhead - The trade-off between model size and compute overhead and reveal there is significant room to reduce the compute-optimal model size with minimal compute overhead.

Papers

1. Language understanding by generative pre-training - Alec et al. openAI

   
2. LLM are few shot learners - scaling LLMs with data is enough to make them few shot.

Models

1. Databricks dolly

   1. Version 1.0
   2. Version 2.0, Huggingface
2. LLaMA
3. Bard
4. StabilityLM

   1. Vicuna

   2. LLaMA

Instructor

1. Training language models to follow instructions with human feedback (using RLHF)
2. Instructor model - "We introduce Instructor👨‍🏫, an instruction-finetuned text embedding model that can generate text embeddings tailored to any task (e.g., classification, retrieval, clustering, text evaluation, etc.) and domains (e.g., science, finance, etc.) by simply providing the task instruction, without any finetuning. Instructor achieves sota on 70 diverse embedding tasks!"

   

Datasets

1. Databricks' 15K QA for Dolly 2.0

Tools

1. Scikit-LLM
2. LangChain

   1. An amazing tutorial in Youtube by Patrick Loeber about

      • LLMs

        • Prompt Templates

        • Chains

        • Agents and Tools

        • Memory

        • Document Loaders

        • Indexes

   2. Langchain in 13 minutes
3. ReAct & LangChain
4. LangFlow, Medium, HuggingFace - is a UI for LangChain, designed with react-flow to provide an effortless way to experiment and prototype flows.
5. PandasAI - PandasAI, asking data Qs using LLMs on Panda's DFs with two code lines. 𝚙𝚊𝚗𝚍𝚊𝚜_𝚊𝚒 = 𝙿𝚊𝚗𝚍𝚊𝚜𝙰𝙸(𝚕𝚕𝚖) & 𝚙𝚊𝚗𝚍𝚊𝚜_𝚊𝚒.𝚛𝚞𝚗(𝚍𝚏, 𝚙𝚛𝚘𝚖𝚙𝚝='𝚆𝚑𝚒𝚌𝚑 𝚊𝚛𝚎 𝚝𝚑𝚎 𝟻 𝚑𝚊𝚙𝚙𝚒𝚎𝚜𝚝 𝚌𝚘𝚞𝚗𝚝𝚛𝚒𝚎𝚜?')
6. LLaMa Index - LlamaIndex (GPT Index) is a project that provides a central interface to connect your LLM's with external data.
7. LLM-foundry - LLM training code for Databricks foundation models using MoasicML
8. Awesome ChatGPT - Curated list of awesome tools, demos, docs for ChatGPT and GPT-3
9. GPT4 All Privacy-oriented software for chatting with large language models that run on your own computer.
10. MinGPT - A minimal PyTorch re-implementation of the OpenAI GPT (Generative Pretrained Transformer) training
11. NanoGPT - The simplest, fastest repository for training/finetuning medium-sized GPTs.
12. Open-sourced codes for MiniGPT-4 and MiniGPT-v2 (https://minigpt-4.github.io, https://minigpt-v2.github.io/)

Guardrails

1. GuardrailsAI

   
2. Safeguarding LLMs with Guardrails

   
3. Databricks GR - Implementing LLM Guardrails for Safe and Responsible Generative AI Deployment on Databricks

Best Practices

1. Best Practices for LLM Evaluation of RAG Applications
2. Announcing MLflow 2.8 LLM-as-a-judge metrics and Best Practices for LLM Evaluation of RAG Applications, Part 2

   

Reinforcement Learning for LLM

1. RLHF: Reinforcement Learning from Human Feedback by Chip Huyen
2. Yoav on RL
3. John Schulman - Reinforcement Learning from Human Feedback: Progress and Challenges

Metrics

1. Understanding ROUGE - a family of metrics that evaluate the performance of a LLM in text summarization, i.e., ROUGE-1, ROUGE-2, ROUGE-L, for unigrams, bi grams, LCS, respectively.

Use Cases

1. Enhancing ChatGPT With Infinite External Memory Using Vector Database and ChatGPT Retrieval Plugin