Calendar | CS 8803

A tentative plan for the schedule (topics, deadlines, etc.) is available here.

Schedule (subject to change as the term progresses)

Aug 26

Training language models to follow instructions with human feedback: PPO/InstructGPT

Aug 28

Direct Preference Optimization: Your Language Model is Secretly a Reward Model: DPO, Unpacking DPO/PPO

Sep 4

Simpo: Simple preference optimization with a reference-free reward: SimPO

Sep 9

Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena: MT-Bench and Chatbot Arena

Sep 11

Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation: CPO

Sep 23

RRHF: Rank Responses to Align Language Models with Human Feedback without tears: RRHF

Sep 25

ORPO: Monolithic Preference Optimization without Reference Model: ORPO

Sep 30

The Llama 3 Herd of Models (Part 1): Llama3 (Part 1)

Oct 2

The Llama 3 Herd of Models (Part 2): Llama3 (Part 2)

Oct 7

VisualWebArena: Evaluating Multimodal Agents on Realistic Visual Web Tasks: VisualWebArena, WebArena

Oct 21

Aya Model: An Instruction Finetuned Open-Access Multilingual Language Model: Aya Model, Aya Dataset

Oct 23

Human-Aware Losses for Alignment (Guest Lecture by Kawin Ethayarajh): KTO

Oct 28

MoMa: Efficient Early-Fusion Pre-training with Mixture of Modality-Aware Experts: MoMa, Switch Transformers

Oct 30

Branch-Train-Merge: Embarrassingly Parallel Training of Expert Language Models: Branch-Train-Merge

Nov 4

LoRA: Low-Rank Adaptation of Large Language Models: LoRA

Nov 6

LESS: Selecting Influential Data for Targeted Instruction Tuning: LESS

Nov 25

QLoRA: Efficient Finetuning of Quantized LLMs: QLoRA